Gcat 2021-Nqm1

Presale: GCAT 2021-NQM1

February 25, 2021

PRIMARY CREDIT ANALYST

Preliminary Ratings Sergey Voznyuk, CFA New York Preliminary Initial interest Preliminary amount Credit enhancement + 1 (212) 438 3010 Class rating(i) Class type rate (%)(ii) ($) (%)(iii) sergey.voznyuk A-1 AAA (sf) Senior Fixed 217,781,000 24.75 @spglobal.com

A-2 AA (sf) Senior Fixed 17,220,000 18.80 SECONDARY CONTACT

A-3 A (sf) Senior Fixed 32,849,000 7.45 Kimball Ng New York M-1 BBB (sf) Mezzanine Fixed 11,142,000 3.60 +1 212-438-2250 B-1 BB (sf) Subordinate Fixed 5,788,000 1.60 kimball.ng @spglobal.com B-2 B (sf) Subordinate Fixed 3,618,000 0.35 SURVEILLANCE CREDIT ANALYST B-3 NR Subordinate Net WAC 1,013,266 0.00 Truc T Bui A-IO-S NR Excess servicing (iv) Notional(v) N/A San Francisco X NR Monthly excess cash (vi) Notional(v) N/A + 1 (415) 371 5065 flow truc.bui @spglobal.com R NR Residual N/A N/A N/A ANALYTICAL MANAGER Note: This presale report is based on information received as of Feb. 25, 2021. The ratings shown are preliminary. Subsequent information may result in the assignment of final ratings that differ from the preliminary ratings. Accordingly, the preliminary ratings should not be construed Vanessa Purwin as evidence of final ratings. This report does not constitute a recommendation to buy, hold, or sell securities. (i)The collateral and structural New York information in this report reflects information received as of Feb. 25, 2021. This information is subject to change, based on the pool as of the + 1 (212) 438 0455 deal cut-off date delivered to the trust on the closing date. The preliminary ratings address our expectation for the ultimate payment of interest and principal. (ii)Interest can be deferred on the classes. Fixed coupons on class A-1, A-2, A-3, M-1, B-1, and B-2 certificates are vanessa.purwin subject to the pool's net WAC. Interest on class B-3 certificates equal pool's net WAC. (iii)This credit enhancement is solely from subordination. @spglobal.com Excess spread also provides credit enhancement. (iv)Excess servicing strip and reimbursable advance. (v)The notional amount equals the aggregate stated principal balance of the loans. (vi)Certain excess amounts from the pool's net WAC over classes with fixed-rate coupons calculated per the pooling and servicing agreement. WAC--Weighted average coupon. NR--Not rated. N/A--Not applicable.

Profile

Expected closing date March 10, 2021.

Deal cut-off date Feb 1, 2021.

Distribution date The 25th of each month, or the next business day, beginning March 2021.

Stated maturity date Jan 25, 2066.

Certificate balance, $289.41 million in aggregate. including unrated classes

www.standardandpoors.com February 25, 2021 1

Profile (cont.)

Collateral types First-lien, fixed- and adjustable-rate, fully amortizing, and interest-only residential mortgage loans primarily secured by single-family residential properties, planned-unit developments, condominiums, cooperatives, and two- to four-family residential properties to both prime and nonprime borrowers. The pool has 535 loans, which are nonqualified or ATR-exempt mortgage loans.

Collateral U.S. residential mortgage loans.

Credit enhancement For each class of preliminary rated certificates, subordination in the form of certificates that are lower in payment priority as well as excess spread that preserves subordination.

ATR--Ability to repay.

Participants

Issuer GCAT 2021-NQM1 Trust.

Sponsor Blue River Mortgage II LLC.

Seller MAT II LLC

Depositor GCAT NQM Depositor II LLC.

Certificate registrar, trustee, and securities Citibank N.A. administrator

Master servicer Nationstar Mortgage LLC.

Servicer NewRez LLC (d/b/a Shellpoint Mortgage Servicing).

Servicing administrator Red Creek Asset Management LLC.

Custodian U.S. Bank N.A.

Originators Arc Home Loan LLC, Quontic Bank, Commerce Home Mortgage LLC, and HomeXpress Mortgage Corp.

Loan data agent DV01 Inc.

D/b/a--Doing business as.

Primary Originators/Initial Loan Sellers

Entity By balance (%) Due diligence (%) Originator ranking

Quontic Bank 31.90 100 N/A

Arc Home Loan LLC 29.83 100 N/A

HomeXpress Mortgage Corp. 28.95 100 N/A

Commerce Home Mortgage LLC 9.32 100 N/A

N/A--Not applicable.

Servicers

S&P Global Ratings' select Entity By balance (%) servicer Operation

NewRez LLC (d/b/a Shellpoint Mortgage 100.00 Yes Primary servicer Servicing)(i)

www.standardandpoors.com February 25, 2021 2

Servicers (cont.)

S&P Global Ratings' select Entity By balance (%) servicer Operation

Nationstar Mortgage LLC 100.00 Yes Master servicer

(i)As of the cut-off date, approximately 19.2% of the pool balance are interim-serviced by BSI Financial Services Inc., LoanCare LLC, or Quontic Bank, and are scheduled to transfer to Shellpoint Mortgage Servicing on or before March 15, 2021. D/b/a--Doing business as.

Rationale

The preliminary ratings assigned to GCAT 2021-NQM1's mortgage pass-through certificates reflect our view of:

- The asset pool's collateral composition (see the Collateral Summary section below);

- The transaction's credit enhancement;

- The transaction's associated structural mechanics;

- The transaction's geographic concentration;

- The transaction's representation and warranty (R&W) framework;

- The mortgage aggregator, Blue River Mortgage II LLC (BRM II); and

- The impact the COVID-19 pandemic will likely have on the performance of the mortgage borrowers in the pool (see "Economic Research: Staying Home For The Holidays," published Dec. 2, 2020) and the liquidity available in the transaction.

A S&P Global Ratings believes there remains high, albeit moderating, uncertainty about the evolution of the coronavirus pandemic and its economic effects. Vaccine production is ramping up and rollouts are gathering pace around the world. Widespread immunization, which will help pave the way for a return to more normal levels of social and economic activity, looks to be achievable by most developed economies by the end of the third quarter. However, some emerging markets may only be able to achieve widespread immunization by year-end or later. We use these assumptions about vaccine timing in assessing the economic and credit implications associated with the pandemic (see our research here: www.spglobal.com/ratings). As the situation evolves, we will update our assumptions and estimates accordingly.

Overview

GCAT 2021-NQM1 is the sixth RMBS transaction rated by S&P Global Ratings under the GCAT shelf. Non-qualified mortgage (non-QM) loans make up 48.3% of the pool, while 51.7% are exempt from the qualified mortgage (QM) rules because they are either investor property loans or loans originated by Quontic Bank (Quontic) or Commerce Home Mortgage LLC (Commerce), which are community development financial institutions. The seller, MAT II LLC, aggregated the loans from four originators via BRM II-affiliated parties. The non-zero weighted average original FICO score of the pool is 742, which is higher than those for prior GCAT transactions. The weighted average original combined loan-to-value (CLTV) of the pool is 69.7%, which is slightly higher than GCAT's prior transactions. (see chart 1).

www.standardandpoors.com February 25, 2021 3

Chart 1

Noteworthy Features

Loans with floating rates indexed to the Secured Overnight Financing Rate (SOFR)

Loans with interest rates indexed to six months SOFR represent approximately 2.3% of the pool balance. We applied stresses to the SOFR rates in accordance with our criteria (see "Methodology To Derive Stressed Interest Rates In Structured Finance," published Oct. 18, 2019).

No loans in forbearance:

On March 27, 2020, the Coronavirus Aid, Relief, and Economic Security (CARES) Act enacted COVID-19 pandemic-related relief for borrowers with government-backed mortgage loans in the form of a temporary forbearance of up to 12 months of scheduled payments. While non-agency loans do not fall under the CARES Act as it relates to this forbearance, servicers have been granting forbearance plans to non-agency borrowers also, typically with some variations to those of the CARES Act (e.g., timeframe, approval requirements, etc.). The updates we made on April 17, 2020, to our mortgage outlook and corresponding archetypal foreclosure frequency levels (see "Guidance: Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later," published April 17, 2020) account for a portion of borrowers entering COVID-19 pandemic-related temporary forbearance plans and their impact to the overall credit quality of collateralized pools. To the extent a securitization pool exhibits growth levels in forbearance over time beyond those otherwise expected, additional adjustments may be applied.

As of the cut-off date, there were no mortgage loans in GCAT 2021-NQM1 where borrowers were

www.standardandpoors.com February 25, 2021 4

granted payment relief as a result of the COVID-19 pandemic; therefore, no additional adjustments were applied.

Servicing

As of the cut-off date, approximately 19.2% of the pool balance are interim-serviced by BSI Financial Inc., LoanCare LLC, or Quontic Bank, and are scheduled to transfer to Shellpoint Mortgage Servicing (Shellpoint) on or before March 15, 2021. Neither interim servicer is required to advance any delinquent interest or principal on the related interim serviced mortgage loans, provided that Shellpoint will be required to make any such advances on the interim serviced mortgage loans during the interim servicing period and after the related servicing transfer date.

Shellpoint must advance delinquent principal and interest (P&I) on any delinquent mortgage loan until it is either greater than 90 days' delinquent (limited P&I advancing) or the P&I advance is deemed unrecoverable by the related servicer. This is a change from prior GCAT transactions where servicers were typically obligated to advance up to 180 days of P&I.

We applied a delinquency stress to test the transaction's liquidity due to the limited nature of P&I advancing and because advancing is not required on a loan while it is in a state of temporary forbearance due to COVID-19-related hardships. We assumed that 35.00% of the cut-off loan balance would be in forbearance (and noncash flowing) for the first six months of the transaction, with any P&I payments related to this delinquent portion coming back to the transaction after all defaults have been passed through to the transaction (approximately 144 months).

Community development financial institution loans

Approximately 41.2% of the loans in pool were originated by Commerce and Quontic, which are designated by the U.S. Department of the Treasury as community development financial institutions (CDFIs). Loans originated by a CDFI are exempt from the ability-to-repay (ATR) rules; therefore, all Commerce- and Quontic-originated loans, including the ones for a primary residence, are ATR exempt. Since the lender did not have to determine the related borrower's ATR for the CDFI loans, these loans could have greater risk than mortgage loans for which the ATR rules applied. Despite this, the CDFI loans' collateral profile is strong, with a weighted average non-zero original FICO score of 744 and an average original CLTV of 65.7%. We also reviewed the underwriting guidelines of the Commerce- and Quontic-originated loans and believe our loan-level LEVELS model accounts for any associated risk posed by the CDFI loans (see the Documentation Type section below for more detail).

Reduced risk retention

Unless an exemption exists, an ABS securitizer is generally required to retain at least 5.0% (the required credit risk) of the aggregate credit risk of the assets collateralizing the securities. The CDFI loans in this transaction fall within the definition of community-focused residential mortgages and are exempt from U.S. risk retention rules. Consequently, because CDFI loans comprise approximately 41.2% of the pool, the transaction's required credit risk retention will be reduced to 2.94% from 5.0%.

Geographic concentration in California

The top five core-based statistical areas (CBSAs) (four in California and one in the New York/New

www.standardandpoors.com February 25, 2021 5

Jersey metropolitan area) account for roughly 49.2% of the aggregate pool. The Los Angeles/Anaheim CBSA accounts for 16.2% of the pool balance and the state of California accounts for 45.1%. We applied a 1.06x adjustment factor to the loss coverages because of this geographic concentration.

Optional Redemption

An optional redemption can occur at the earlier of the February 2024 distribution date or when the loans' aggregate stated principal balance is less than or equal to 30.0% as of the cut-off date. This feature remains the same as in the GCAT 2020-NQM2 transaction.

Optional purchase of delinquent loans

The depositor or its designee, at its option, may purchase any mortgage loan that is 90 or more days Mortgage Bankers Assn. (MBA) delinquent (including any loans subject to a forbearance plan related to the COVID-19 pandemic that are 90 or more days MBA delinquent following the end of the forbearance period), or any real estate-owned (REO) property acquired in respect of a mortgage loan from the issuing trust at the purchase price, which covers principal at par, accrued interest, pre-closing deferred amounts and relevant expenses. The aggregate-stated principal balance of such mortgage loans and REO properties purchased by the sponsor cannot exceed 7.5% of the aggregate stated principal balance of the mortgage loans as of the cut-off date.

Collateral Summary

GCAT 2021-NQM1's assets consist of fixed- and adjustable-rate, and interest-only non-QM loans secured by first liens. The asset pool consists of 535 mortgage loans with a principal balance of approximately $289.41 million as of the cut-off date.

The collateral pool, from a credit perspective, is weaker than the S&P Global Ratings' archetypal prime pool, but it is generally in line with our expectations of a nonprime residential mortgage pool (see table 1). The pool's 'AAA' loss coverage requirement was determined to be 17.75%. In our analysis, we considered the following mortgage loan characteristics to be weaker:

- A significant number of non-QM loans;

- Alternative income documentation on loans;

- Property-focused investor loans;

- Loan type (adjustable-rate mortgage loans and interest-only term features);

- Loan terms (40-year term loans);

- Geographic concentration;

- Property type (elevated two-to-four family homes);

- Occupancy status: second home or investor property;

- Loan purpose: cash-out refinances; and

- Loan type: adjustable-rate mortgage loans or interest-only term features.

The mortgage loans consist of adjustable-rate, fully amortizing mortgage loans (19.7% of the pool balance); fixed-rate, fully amortizing mortgage loans (66.7%); adjustable-rate, interest-only

www.standardandpoors.com February 25, 2021 6

(5.0%); and fixed-rate, interest-only loans (8.7%). The weighted average seasoning for the pool is approximately six months from the loans' first payment date to the cut-off date.

The weighted average used FICO score of 742 is higher than prior GCAT transactions and the weighted average original CLTV of 69.7% is slightly higher than GCAT's prior transactions. This combination overall resulted in a FICO/LTV adjustment factor of 0.89x in our model, which is lower than the prior two GCAT transactions.

Compared to prior GCAT transactions, GCAT 2021-NQM1 consists of fewer property-focused business purpose loans, cash-out refinance loans, and adjustable-rate loans. GCAT 2021-NQM1 also has a lower DTI ratio than previous transactions. The strengths compared to prior GCAT transactions are partly offset by a higher concentration of loans with income verified using alternative methods such as bank statements (see chart 1).

The weighted average used FICO score includes certain S&P Global Ratings assumptions (see table 2 for a breakdown of the pool by borrower FICO score). In the pool, there are 23 loans to foreign borrowers (2.7% of the pool balance), of which six are to a borrower without a FICO score. We assessed this loan in our credit analysis using a FICO score of 699, which is approximately the mortgage pool's average original FICO score minus one standard deviation. For all loans to foreign borrowers, we applied a 1.5x multiple to the foreclosure frequencies.

Mortgage loans backed by properties that are primary residences make up approximately 73.5% of the pool balance. The mortgage loans are secured by first liens on single-family residences (59.0% of the pool balance), planned-unit developments (PUDs; 24.3%), condominiums (6.3%), co-op (0.60%) and two- to four-family homes (9.9%) (see table 1).

Table 1

Collateral Characteristics(i)

GCAT GCAT GCAT GCAT GCAT GCAT Archetypal 2021-NQM1 2020-NQM2 2020-NQM1 2019-NQM3 2019-NQM2 2019-NQM1 pool(ii)

Closing pool balance (mil. 289.4 226.9 335.4 314.6 402.9 393.0 N/A $)

Closing loan count (no.) 535 353 621 820 1056 876 N/A

Avg. loan balance ($) 540,956 642,773 540,140 383,651 381,604 448,672 N/A

WA original CLTV (%) 69.7 69.7 69.5 65.4 65.5 66.2 75.0

WA current CLTV (%) 69.1 69.4 69.0 64.4 63.4 65.2 75.0

WA FICO(iii) 742 734 720 735 731 718 725

WA current rate (%) 5.7 5.4 5.8 5.7 5.9 6.1 N/A

WA original term (mos.) 362 373 381 368 368 361 360

WA seasoning (mos.) 6 5 3 6 9 9 0-6

WA debt-to-income (%) 33.3 37.9 36.5 35.9 36.5 37.3 36.0

WA DSCR (non-zero) 1.2 1.1 1.1 1.2 1.2 1.2 N/A

Owner occupied (%) 73.5 68.0 70.0 62.8 64.2 76.3 100.0

Single-family (including 83.2 77.9 80.0 78.5 77.5 72.3 100.0 unattached and attached PUD) (%)

Adjustable-rate loans (%) 24.6 42.1 56.1 85.2 83.5 70.0 0.0

Loans with IO payments 13.7 37.1 34.2 10.7 10.0 4.5 0.0 (%)

www.standardandpoors.com February 25, 2021 7

Table 1

Collateral Characteristics(i) (cont.)

GCAT GCAT GCAT GCAT GCAT GCAT Archetypal 2021-NQM1 2020-NQM2 2020-NQM1 2019-NQM3 2019-NQM2 2019-NQM1 pool(ii)

Purchase (%) 68.4 32.3 41.9 62.8 57.4 60.8 100.0

Cash-out refinancing (%) 25.4 41.2 42.4 28.0 29.7 32.8 0.0

Full documentation (%) 20.0 29.2 29.7 51.1 36.7 34.6 100.0

Alternative/bank 68.9 43.2 45.5 39.2 42.1 52.4 0.0 statement documentation (%)

Other/asset 11.1 27.6 24.8 9.6 21.2 13.0 0.0 depletion/DSCR documentation (%)

Investor property % 22.2 29.8 28.5 22.5 33.9 20.3 0.0

Self-employed borrowers 76.2 58.9 54.0 43.9 53.3 59.7 0.0 (%)

Loans with co-borrowers 27.9 36.8 33.2 21.3 22.9 26.3 0.0 (%)

Loans to borrowers with 3.4 6.5 5.5 1.6 3.6 2.4 N/A multiple mortgages (%)(iv)

Loans to foreign 2.7 1.1 3.1 3.7 3.8 4.1 0.0 borrowers (%)(foreign national and non-permanent resident aliens)

Modified loans (%)(v) 0.0 0.0 0.0 0.0 0.0 0.0 0.0

PCEs (%)(v) 0.3 0.5 0.9 0.3 0.5 0.8 0.0

Current (%) (viii) 98.1 100.0 100.0 100.0 100.0 100.0 100.0

30+ day delinquent (%) 1.9(ix) 0.0 0.0 0.0 0.0 0.0 0.0 (viii)

Length of P&I advancing 3 6 6 6 6 6 Full (mos.)(vi)

Pool-level adjustments (multiplicative factors)

Geographic 1.06 1.14 1.09 1.22 1.10 1.18 1.00 concentration

Mortgage operational 1.05 1.05 1.05 1.05 1.05 1.05 1.00 assessment

Representations and 1.10 1.10 1.10 1.10 1.10 1.10 1.00 warranties

Other (i.e. loan 1.00 1.01 1.01 1.01 1.01 1.01 1.00 modification/PCE/due diligence)

Loans in 1.00 1.15 N/A N/A N/A N/A N/A forbearance/deferred payements related to COVID-19

Combined pool-level 1.22 1.53 1.27 1.41 1.28 1.38 1.00 adjustments(vii)

www.standardandpoors.com February 25, 2021 8

Table 1

Collateral Characteristics(i) (cont.)

GCAT GCAT GCAT GCAT GCAT GCAT Archetypal 2021-NQM1 2020-NQM2 2020-NQM1 2019-NQM3 2019-NQM2 2019-NQM1 pool(ii)

Loss estimation

'AAA' loss coverage 17.75 28.35 28.30 18.75 20.50 23.50 7.50 (%)

'AAA' foreclosure 37.52 54.31 55.89 37.01 43.08 47.34 15.00 frequency (%)

'AAA' loss severity (%) 47.31 52.2 50.64 50.66 47.59 49.64 50.00

'BBB' loss coverage 5.55 9.70 9.15 5.35 6.00 6.75 1.92 (%)

'BBB' foreclosure 20.30 30.66 29.41 18.22 21.76 24.11 6.41 frequency (%)

'BBB' loss severity (%) 27.34 31.64 31.11 29.36 27.57 28.00 30.00

'B' loss coverage (%) 1.60 3.15 2.50 1.40 1.50 1.65 0.65

'B' foreclosure 8.36 13.74 10.80 6.64 7.70 8.48 3.25 frequency (%)

'B' loss severity (%) 19.14 22.93 23.15 5.31 19.48 6.78 20.00

(i)Information as of cut-off date for each transaction. (ii)As defined in our Feb. 22, 2018, criteria article. (iii)FICO reflects the most recent scores obtained. We assumed a FICO of 699 for borrowers missing FICO which was the pool average FICO minus one standard deviation of the FICO (iv)Limited to borrowers who have multiple mortgage loans or properties included in the securitized pool. (v)Limited to modified and PCE loans considered in our analysis. (vi)Months of P&I advancing on a delinquent mortgage loan to the extent such advances are deemed recoverable. (vii)Combined pool-level adjustments are the product of each pool-level adjustment listed above. (viii) Loans in forbearance are treated as current and included in the above the model forbearance adjustment. (ix)Twelve loans were 30 days MBA delinquent as of the cut-off date; however, provided that 11 loans subsequently made the delinquent payment and the sponsor will remove the remaining one loan from the pool prior to the closing date if the related mortgagors do not make the related delinquent payments by Feb. 26, 2021. WA--Weighted average. CLTV--Combined loan-to-value ratio. DSCR--debt service coverage ratio. PUD--planned-unit development. IO--Interest only. PCE--Prior credit event. P&I--Principal and interest. N/A--Not applicable.

Table 2

Updated Credit Score Statistics

FICO score Current balance (%) No. of loans Average current balance ($000s)

750+ 43.65 229 551.6

725-749 22.93 112 592.5

700-724 17.57 100 508.5

675-699 11.59 61 549.9

650-674 3.31 23 416.0

625-649 0.79 8 285.1

600-624 0.07 1 203.8

575-599 ------

550-574 ------

Below 550 0.10 1 284.3

Total 100.0 535 541.0

www.standardandpoors.com February 25, 2021 9

Transaction Structure

Chart 2 shows an overview of the transaction's structure.

Chart 2

The transaction is structured as a true sale of the receivables from the seller, MAT II LLC, to the depositor, GCAT NQM Depositor II LLC, and a transfer from the depositor to the issuing trust, GCAT 2021-NQM1 Trust. The issuing trust transfers the certificates to the depositor, and the depositor sells them to the initial purchaser, who then sells them to third-party investors. The depositor sells the non-offered certificates, as well as the certificates required to be held to satisfy the risk retention rules, to either the sponsor, BRM II, or a majority-owned affiliate.

In rating this transaction, S&P Global Ratings will review the legal matters it believes are relevant to its analysis, as outlined in its criteria.

www.standardandpoors.com February 25, 2021 10

Strengths And Weaknesses

We believe the following characteristics strengthen the GCAT 2021-NQM1 transaction:

- The mortgage pool generally consists of loans to borrowers with higher FICOs and significant home equity relative to our archetypes, as demonstrated by the pool's weighted average used FICO of 742 and original combined LTV ratio of 69.7%.

- The third-party due diligence providers--Clayton Services LLC, Consolidated Analytics LLC, and Digital Risk LLC (all of which are on our list of reviewed providers)--performed due diligence on 100% of the pool's loans. Their reviews encompassed credit (underwriting) compliance, property valuations, regulatory compliance, and data quality.

- The class A-1, A-2, and A-3 certificates (the senior classes) benefit from a credit support floor in which no principal is paid to the mezzanine and subordinate classes until the senior classes are retired. Additionally, principal is paid sequentially among the senior classes in periods when either the cumulative loss or the delinquency trigger has failed, further protecting the more-senior classes.

We believe the following factors weaken this transaction:

- Income on certain mortgage loans (68.9% of the pool balance) was verified using alternative methods, such as personal or business bank statements, or certified public accountant (CPA)-prepared profit and loss (P&L) statements. We consider income verification using those alternative methods to be a weaker standard than "full" documentation of income. Consequently, we increased our loss coverages for these loans by applying an adjustment to the foreclosure frequencies. We applied adjustment factors of 2.00x and 1.75x to the foreclosure frequencies for loans using 12-23 months or 24 months or more of income verification, respectively.

- About 16.8% of the loans are on property types other than single-family homes and PUDs, such as condominiums and two-to-four family houses. We applied an adjustment to their foreclosure frequencies to account for this risk.

- The loan purpose for 25.4% of the loans by balance is cash-out refinance.

- Non-QM loans, which have an increased risk of ATR challenges and associated loan losses, make up 48.3% of the pool. We applied an adjustment to loss severities to account for this risk, which increased our loss coverage estimates for each rating category.

- A portion of the mortgage loans (15.9% of the pool balance) were made to borrowers with FICO scores below 700 (including our assumption of a 699 FICO score for borrowers who lack a FICO score). The mortgage pool's loss estimate was increased to account for the increased default risk of these loans.

- The respective originators make the R&Ws with the sponsor, backstopping the originators' repurchases in the event of an insolvency of the originator, which is a strength. However, the R&W framework is somewhat weak because the controlling holder (the majority owner of the class X certificates and, initially, an affiliate of the sponsor) is not required to test for breaches on all loans that go delinquent or sustain a realized loss, although it has the option. However, the controlling holder is obliged to review all ATR notice (a violation of the ATR rules raised) loans that suffer a realized loss; and it may have conflicting interests vis-à-vis the certificateholders. The risks and concerns with the framework are somewhat mitigated by third parties performing 100% due diligence on the loans, and the sponsor (through a

www.standardandpoors.com February 25, 2021 11

majority-owned affiliate of the sponsor) holding the first loss piece and retaining risk via a material net economic interest of not less than 2.94% of the capital structure. Consequently, we applied an R&W adjustment that increased our loss expectations at all rating categories by a 1.10x factor.

Credit Analysis And Assumptions

Our analysis of the GCAT 2021-NQM1 collateral pool considered a number of factors, including certain loan-level characteristics detailed below.

Documentation type

The sponsor guidelines allow income verification using paystubs, W-2s/W-2 equivalents, tax returns, CPA-prepared P&L statements, personal or business bank statements, or a letter from a CPA or an accountant documenting income. The sponsor also offers asset depletion programs that consider borrower accumulated assets in calculating income to qualify the borrower. Cash flow investor loans are underwritten, in part, to a DSCR, which is calculated using actual or estimated rent.

Table 3

Documentation Type (Income Verification Type/ Length)

Foreclosure 'AAA' foreclosure Loan Alternative income frequency frequency without count Current verification length adjustment factors pool adjustment (no.) balance (%) (WA # of months) (x) factors (%)

Full documentation

Appendix Q/qualified ------1.00 -- mortgage

Full (24+ months) 39 10.8 -- 1.00 23.0 excluding WVOE

Full (24+ months) 28 3.1 -- 1.00 16.0 WVOE

Full (12-23 months) 3 0.6 -- 1.25 21.4 excluding WVOE

Full (12-23 months) 38 5.1 -- 1.25 23.4 WVOE

Full (1-11 months) ------1.50 -- excluding WVOE

Full (1-11 months) 4 0.4 -- 1.50 28.5 WVOE

Alternative documentation

24+ months (primary source)

Business bank 82 18.8 24.1 1.75 30.2 statements

Personal bank 18 4.0 24.1 1.75 33.0 statements

www.standardandpoors.com February 25, 2021 12

Table 3

Documentation Type (Income Verification Type/ Length) (cont.)

Personal and ------1.75 -- business bank statements(i)

P&L statements(ii) 3 0.5 24.0 1.75 72.0

Additional alternative ------1.75 -- CPA letter/FN 24+

12-23 months (primary source)

Business bank 70 18.7 12.2 2.00 33.9 statements

Personal bank 18 4.7 12.1 2.00 44.4 statements

Personal and 1 0.1 12.0 2.00 100.0 business bank statements(i)

P&L statements(ii) 141 22.2 12.0 2.00 36.7

Additional alternative ------2.00 -- CPA letter/FN 24+

1-11 months (primary source)

Business bank ------2.25 -- statements

Personal bank ------2.25 -- statements

P&L statements ------2.25 --

Additional alternative ------2.25 -- CPA letter/FN 24+

Other documentation

Other (DSCR) 78 8.1 -- 3.15-6.00 51.9

Other (applied 0.00 2 0.7 -- 6.00 90.8 DSCR)

Other (asset 10 2.3 -- 3.00 28.5 depletion)

(i)Both account types provided. (ii)The documentation source may include other secondary documentation types such as a CPA letter or supporting bank statements. A majority of the P&L statements were CPA prepared. WVOE--Written verification of employment/employer letter. WA--Weighted average. P&L--Profit and loss. FN--Foreign national program.

For 112 loans (approximately 20.0% of the pool balance), traditional (full) documentation was used for fully verifying and calculating the borrowers' qualifying income (e.g., written verification of employment, pay stubs, W-2s, personal and business tax returns, and IRS transcripts). We applied a documentation type adjustment factor ranging from 1.00x to 1.50x, depending on the length of the income verification, for these loans.

We classified all loans to borrowers that used nontraditional sources of income documentation, such as bank statements (business or personal) or CPA-prepared P&L statements documenting income as alternative documentation loans. Alternative documentation was used on 333 mortgage loans (68.9% of the pool balance), with most borrowers using 12-23 months of income verification. We view income verification using alternative documentation to be a weaker standard

www.standardandpoors.com February 25, 2021 13

than full documentation of income. Consequently, we increased our loss coverages for these loans by an adjustment factor ranging from 1.75x to 2.00x.

Seventy-eight loans in the pool (8.1% by pool balance) were underwritten to a lending program that considers investment property cash flow. We classified these loans as "other" documentation loans with a debt service coverage ratio (DSCR) flag and applied a 3.15x-6.00x adjustment factor to the foreclosure frequencies based on the provided DSCR calculation. The DSCR calculations provided by the issuer for these loans ranged from 0.68 to 5.07.

We classified two investment property business purpose loans (0.7% by balance) that were primarily underwritten to FICO scores and LTV ratios (rather than DSCR) as "other" documentation type. We evaluated these non-DSCR loans as if they were DSCR loans with a DSCR of zero; our loss model applied a 6.00x adjustment factor to the foreclosure frequencies for these non-DSCR loans.

Ten loans in the pool (2.3% of the pool balance) were underwritten, in full or in part, by a lending program that considers accumulated assets rather than a verified income stream. We classified these loans as other loans and applied a 3.00x adjustment to the foreclosure frequencies.

Of the 249 Commerce- and Quontic-originated loans (which are ATR exempt because they are CDFIs), we classified 72 as full, 160 as alternative, and the remaining 17 as other documentation type loans. Both Commerce and Quontic guidelines generally specify using a written verification of employment or the more commonly used W-2/tax statements to document a wage earner's income. As such, we treated these loans as "full" documentation type loans. For self-employed borrowers, the Quontic guidelines generally specify using a CPA-prepared 12-month P&L statement for income documentation, while Commerce's guidelines specify two years of tax returns (business or personal), in addition to P&L statements; we treated these loans as "alternative" documentation type loans.

Prior credit event classification and analysis

We focused primarily on prior bankruptcy, foreclosure, short sale, and deed-in-lieu events (24 months from the cut-off date for bankruptcy discharges or dismissals and 36 months from the cut-off date for housing-related events). For loans to borrowers with more seasoned PCEs, we believe the risks associated with those PCEs are appropriately reflected in the updated FICO. Only three loans (0.3% of the pool balance) fell into the PCE category, as defined in our criteria. We applied a 1.00x pool-level PCE related loss coverage adjustment factor.

QM and ATR standards

The CFPB issued final regulations for mortgage loans with applications submitted on or after Jan. 10, 2014, specifying the standards for a QM. Based on the designation provided by the sponsor, about 48.3% of the loans (by balance) are categorized as non-QM/compliant (see table 4).

Table 4

Qualified Mortgage Breakout

QM status Pool balance ($) % by pool balance Loan count (no.) Weighted average FICO score

QM/non-HPML ------

QM/HPML ------

Non-QM/compliant 139,737,489 48.3 200 739

www.standardandpoors.com February 25, 2021 14

Table 4

Qualified Mortgage Breakout (cont.)

QM status Pool balance ($) % by pool balance Loan count (no.) Weighted average FICO score

Not covered/exempt 149,673,778 51.7 335 744

QM--Qualified mortgage. HPML--Higher-priced mortgage loan.

Under the ATR rule, the originator and any assignee are jointly and separately liable for certain damages that may be incurred from noncompliance with the rule (see Appendix I of "Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later," published Feb. 22, 2018). For each loan subject to the rule, we applied an adjustment to loss severities to account for this risk. The data the issuer provided to S&P Global Ratings, including additional fields that validate the loan's QM designation, were reviewed by the due diligence firms under the third-party due diligence firms' scope to verify that documentation exists to support the QM designation. Also, in conjunction with our aggregator review, we concluded that BRM II's processes address the ATR risks.

Servicer advancing obligations

For any loan that is not in forbearance, the servicer, NewRez LLC (doing business as Shellpoint), must advance delinquent P&I on any delinquent mortgage loan until it is either greater than 90 days' delinquent (limited P&I advancing) or the P&I advance is deemed unrecoverable by the related servicer. In the event that the servicer fails to advance P&I, the master servicer, Nationstar Mortgage LLC, is obligated to make those advances unless the master servicer determines that the P&I advance is non-recoverable. To the extent that the servicer and the master servicer fail to make a required P&I advance, the securities administrator will be obligated to make that required P&I advance.

Approximately 19.2% of the pool balance as of the cut-off date are interim serviced by BSI Financial Services Inc., LoanCare LLC, or Quontic Bank, and are scheduled to transfer to Shellpoint on or before March 15, 2021. Neither interim servicer is required to advance any delinquent interest or principal on the related interim serviced mortgage loans, provided Shellpoint will be required to make any such advances on the interim serviced mortgage loans during the interim servicing period and after the related servicing transfer date.

Unlike P&I advances, the servicers must make advances of delinquent taxes and insurance (and other property preservation advances) on any delinquent mortgage loan until the related property is liquidated or the servicer deems the advance to be unrecoverable. We incorporated the limited P&I advancing into our loss severity calculations.

Borrowers with multiple loans

We did not make an additional adjustment to the loss coverage and the tail risk analysis due to borrowers with multiple loans in the transaction because only 16 borrowers have multiple loans. These loans combined represent 3.4% of the pool balance.

www.standardandpoors.com February 25, 2021 15

Structural Features

GCAT 2021-NQM1 structure has a mix of pro rata and sequential features. Principal is paid pro rata among the senior classes (subject to passing cumulative loss and delinquency trigger tests) and then sequentially to the mezzanine and subordinate classes. In the periods when the cumulative loss or delinquency trigger test fails, principal is first used to pay any unpaid interest and interest carryforward amounts (to the extent not paid after allocation of the interest remittance amount) sequentially to the class A-1 and the class A-2 certificates, then pay principal sequentially to classes A-1 and A-2; and then pay any unpaid interest and interest carryforward amounts to the classes, followed by principal to that class until reduced to zero--sequentially to the A-3, M-1, B-1, B-2, and B-3 certificates in that order, with both interest and principal paid to a class before payments to the next class (see table 6).

Since the class A-1, A-2, and A-3 certificates can receive principal pro rata, the amount of protection to the class A-1 and A-2 certificates can decline over time. In our analysis, the delinquency or cumulative loss trigger may help protect the more-senior classes by allowing the payment mechanism to switch to sequential earlier, thus preserving subordination and requiring less upfront credit enhancement.

The transaction also uses monthly excess cash flow to cover the current period's realized losses and to reimburse any previously applied realized loss amounts.

The securities administrator will make monthly distributions of interest from the interest remittances and principal from principal remittances (see tables 5-7).

Table 5

Interest Payment Waterfall

Priority Payment

1 Interest and interest carryforward amounts(i) sequentially to the class A-1, A-2, A-3, M-1, B-1, B-2, and B-3 certificates.

2 Any remaining amount paid as part of monthly excess cash flow.

(i)Interest carryforward amount are deferred interest payments that accrue interest at the lower of the respective coupon and the net WAC rate. Our ratings address the expectation for full payment of all interest and interest carryforward amounts by the final maturity date. WAC--Weighted average coupon.

Table 6

Principal Payment Waterfall

Priority Payment

If the delinquency and cumulative loss trigger tests pass

1 Unpaid interest and interest carryforward amounts sequentially to the class A-1, A-2, and A-3 certificates.

2 Principal pro rata to the class A-1, A-2, and A-3 certificates.

3 Unpaid interest and interest carryforward amounts to the class M-1 certificates.

4 Principal to the class M-1 certificates.

5 Unpaid interest and interest carryforward amounts to the class B-1 certificates.

6 Principal to the class B-1 certificates.

7 Unpaid interest and interest carryforward amounts to the class B-2 certificates.

www.standardandpoors.com February 25, 2021 16

Table 6

Principal Payment Waterfall (cont.)

Priority Payment

8 Principal to the class B-2 certificates.

9 Unpaid interest and interest carryforward amounts to the class B-3 certificates.

10 Principal to the class B-3 certificates.

11 Any remaining amount paid as part of monthly excess cash flow.

If the delinquency or cumulative loss trigger tests fail

1 Unpaid interest and interest carryforward amounts to the class A-1 certificates.

2 Unpaid interest and interest carryforward amounts to the class A-2 certificates.

2 Principal to the class A-1 certificates.

4 Principal to the class A-2 certificates.

5 Unpaid interest and interest carryforward amounts to the class A-3 certificates.

6 Principal to the class A-3 certificates.

7 Unpaid interest and interest carryforward amounts to the class M-1 certificates.

8 Principal to the class M-1 certificates.

9 Unpaid interest and interest carryforward amounts to the class B-1 certificates.

10 Principal to the class B-1 certificates.

11 Unpaid interest and interest carryforward amounts to the class B-2 certificates.

12 Principal to the class B-2 certificates.

13 Unpaid interest and interest carryforward amounts to the class B-3 certificates.

14 Principal to the class B-3 certificates.

15 Any remaining amount paid as part of monthly excess cash flow.

Table 7

Monthly Excess Cash Flow Waterfall

Priority Payment

1 Sequentially to the class A-1, A-2, A-3, M-1, B-1, B-2, and B-3 certificates up to the realized loss amount for the current period until their respective class certificate balances are reduced to zero.

2 Sequentially, up to the cumulative applied realized loss amount to the class A-1, A-2, A-3, M-1, B-1, B-2, and B-3 certificates until their respective certificate balances are reduced to zero.

3 To the cap carryover reserve account, from monthly excess amount otherwise distributable to the class X certificates; the aggregate cap carryover amount(i) for classes A-1, A-2, A-3, M-1, B-1, and B-2 for such distribution date; and, from the cap carryover reserve account, any unpaid cap carryover amounts sequentially to the class A-1, A-2, A-3, M-1, B-1, and B-2 certificates.

4 Sequentially, any amounts due to the class X certificates, then reimburse ETE amounts that exceed annual caps, pro rata, to the relevant transaction parties, and then any remaining amounts to the class R certificates.

(i)The cap carryover amount is the positive difference between the interest that would have accrued at the coupon rate (without regard to the pool's net WAC rate) and what was actually due based upon the net WAC rate. Any prior unpaid cap carryover amounts also accrue at the coupon rate without regard to the net WAC rate. Our preliminary ratings do not address the payment of cap carryover amounts. PSA--Pooling and servicing agreement. WAC--Weighted average coupon. ETE--Extraordinary trust expenses

The interest remittance amount includes the interest collected from, or advanced on behalf of,

www.standardandpoors.com February 25, 2021 17

borrowers (including interest payments that accompany prepayments, any compensating interest and interest portions of liquidation proceeds [minus expenses], subsequent recoveries, termination prices, redemption prices, and repurchase amounts) minus servicing, master servicing, trustee, loan-data agent, and custodial fees, as well as the servicer advance reimbursements permitted under the pooling and servicing agreement, reimbursable expenses incurred by the controlling holder, and extraordinary expenses, which are generally capped at $300,000 annually. Although the extraordinary expenses are passed through as reduced contractual interest due to certificateholders, we ran these expenses at their capped amounts (as described in the Interest Stresses section below). We also considered the extraordinary expenses when analyzing projected interest reduction amounts (as described in the Imputed Promises Analysis section below).

Principal remittance amounts include the principal collected from or advanced on behalf of borrowers (including prepayments, principal portions of liquidation proceeds [minus expenses], subsequent recoveries, termination price, redemption price, and repurchase amounts) minus fees, including extraordinary trust expenses that could not be paid from interest collections.

Interest on the class A-1, A-2, A-3, M-1, B-1, and B-2 certificates is based on the lower of the coupon rate on the certificates and the net weighted average coupon (WAC) rate. Interest on class B-3 is the net WAC rate. The net WAC rate is defined as the weighted average of the mortgage interest rates of the loans, net of fees and extraordinary expenses, weighted based on the loans' stated principal balances. Our preliminary ratings address the lower of these two rates.

Under the transaction documents, the issuer can defer interest payments on these securities. A failure to pay the interest amounts due on the securities will result in the interest being deferred. Deferred interest (the interest carryforward amount) accrues interest at the lower of the fixed rate and net WAC rate for each class. Our preliminary ratings address the expectation for P&I payments (including interest carryforward amounts) by the certificates' final maturity date.

However, our preliminary ratings do not address the expectation for payment of the cap carryover amounts (i.e., the difference between the coupon and the net WAC where the coupon exceeds the net WAC). These amounts are subordinated in the payment priority. In our view, neither the initial coupons on the certificates nor the initial net WAC rates are de minimis, and the nonpayment of the cap carryover amount is not considered an event of default under the transaction documents. Therefore, in line with our criteria for imputed promises, we do not consider whether these cap carryover amounts are paid in our cash flow analysis.

In GCAT 2021-NQM1, the mezzanine and subordinate certificates are paid principal sequentially after all senior certificates have been paid. Unlike the credit enhancement seen in shifting-interest RMBS structures, which may deplete due to scheduled and prepaid principal paid to the subordinate classes, the credit enhancement in GCAT 2021-NQM1 is not depleted because principal payments are not made to these classes unless they are the most senior class outstanding.

Although principal is paid pro rata among the senior classes from the start and there is no defined credit enhancement floor that would switch the senior classes' payment priority to sequential, we are comfortable that the transaction is adequately enhanced for the assigned preliminary ratings, taking into account any tail risk considerations given that the transaction starts with 7.45% enhancement for the senior classes, which then grows as a percentage of the current balance as they get paid down (see the Large Loan And Tail Risk Considerations section). Additionally, the delinquency trigger and cumulative loss rate trigger (see tables 8 and 9) protect the more senior classes in tail risk situations if defaults were to increase much later in the transaction's life (a back-ended default curve) by switching the payment priority among the senior classes to

www.standardandpoors.com February 25, 2021 18

sequential.

Table 8

Cumulative Loss Trigger Event

Distribution date occurring in the following Applied realized loss amounts since the closing date (as a % of the periods cut-off date pool balance)

March 2021 through February 2024 2.00

March 2024 through February 2025 3.00

March 2025 through February 2026 4.00

March 2026 and thereafter 7.00

Table 9

Delinquency Trigger Event

Distribution date occurring in the Six-month average of 60+ day delinq. plus loans modified in past 12 months following periods including any deferrals (as a % of the current pool balance)

March 2021 through February 2024 20.00

March 2024 through February 2026 25.00

March 2026 and thereafter 30.00

Delinq.--Delinquent.

If the aggregate class balance of the certificates exceeds the pool balance, the resulting excess (the applied realized loss amount) is applied in sequential order to the class B-3, B-2, B-1, M-1, A-3, A-2, and A-1 certificates until each class' principal balance has been reduced to zero.

If the pool balance exceeds the aggregate class balance of the certificates (after the allocation of principal payments and monthly excess cash flow to pay down the certificates), the balances of the class A-1, A-2, A-3, M-1, B-1, B-2, and B-3 certificates will be written up sequentially in that order to the aggregate amount of applied realized losses previously allocated.

Geographic Concentration

S&P Global Ratings analyzes the pool's geographic concentration risk based on the concentrations of loans in each of the CBSAs defined by the U.S. Office of Management and Budget (see Appendix II of "Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later," published Feb. 22, 2018). In this transaction, the top five CBSAs account for 49.2% of the aggregate pool. We applied a geographic concentration adjustment factor of 1.06x to our base loss coverage estimate (see table 10).

Table 10

Geographic Concentration

CBSA code(i) CBSA State % by balance

31084 Los Angeles-Long Beach-Glendale (Metropolitan Division) California 16.18

35614 New York-Jersey City-White Plains (Metropolitan Division) New York 13.94

11244 Anaheim-Santa Ana-Irvine (Metropolitan Division California 8.88

35004 Nassau County-Suffolk County (Metropolitan Division) New York 6.00

www.standardandpoors.com February 25, 2021 19

Table 10

Geographic Concentration (cont.)

CBSA code(i) CBSA State % by balance

41740 San Diego-Chula Vista-Carlsbad California 4.49

Top five -- -- 49.23

(i)CBSA code refers to the metropolitan division code, if available. CBSA--Core-based statistical area (includes metropolitan statistical areas and metropolitan divisions where defined, as well as micropolitan statistical areas).

Large Loans And Tail Risk Considerations

As the number of loans in the transaction decreases, the effect of a single loan's losses becomes greater. If conditional prepayment rates are slow and collateral pool losses are not realized until later in a transaction's life (back-loaded losses), pro rata pay mechanisms can then leave the senior certificates exposed to event risk later in the transaction's life (for more information on tail risk in RMBS transactions, see "Older RMBS Transactions Face Increased Tail Risk As Their Pools Shrink," published Aug. 9, 2012).

To mitigate this risk, certain transactions provide for a credit enhancement floor, specifying principal payments not be made to mezzanine or subordinate classes if the credit support available to the senior classes falls below a threshold. GCAT 2021-NQM1 does not explicitly provide a credit enhancement floor. However, due to the sequential payment mechanism to the mezzanine and subordinate classes, which make up 7.45% of the capital structure, the preliminary 'AAA (sf)', 'AA (sf)', and 'A (sf)' rated classes effectively have a floor of 7.45% initially. Although over time, subordination can be depleted due to realized losses, the effective floor to the more-senior classes can increase when losses or delinquencies go over certain thresholds and trip the cumulative loss or the delinquency triggers, making the payment priority fully sequential. In our stress rating scenarios, either the triggers tripped, and the transaction went fully sequential or when the senior classes were paid pro rata, the subordination to the senior classes was adequate.

To analyze the appropriateness of this effective credit enhancement floor, we use an approach outlined in "Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later," published Feb. 22, 2018. Per this approach, instead of focusing on the largest loans by balance at issuance, we risk-weight the loans in the transaction by focusing on those loans with the largest expected loss exposure, assuming default.

After considering the credit enhancement provided in the transaction in conjunction with the cumulative loss and delinquency trigger definitions and the certificates' expected paydown, we believe the rated senior certificates are sufficiently protected from tail risk as the transaction seasons.

Mortgage Operational Assessment (MOA)

Angelo, Gordon and Co. L.P. (Angelo Gordon) was founded in 1988 and the structured credit platform at Angelo Gordon began operations in 2008. Blue River Mortgage TRS (BRM) was formed in 2017 as a privately held real estate investment trust that is managed by Angelo Gordon and owned by certain funds managed by Angelo Gordon. Subsequently in July 2020, BRM II was formed as an additional vehicle to acquire non-QM mortgage loans via the same policies,

www.standardandpoors.com February 25, 2021 20

procedures, controls, and staffing of BRM. For the purposes of this transaction we incorporated our MOA of BRM (detailed below) to BRM II.

Given its management and ownership structure, BRM benefits from various resources provided by Angelo Gordon, such as employees, technology, oversight, and the general expertise of a firm with over 30 years of experience and more than $43 billion in assets under management (AUM). BRM acquires non-QM mortgages through bulk and flow purchases with the mortgage servicing rights typically released. NewRez LLC (doing business as Shellpoint), an S&P Global Ratings select servicer; AmWest Funding Corp.; Metro City Bank; and Royal Business Bank, currently service the loans in the BRM portfolio.

Since December 2017, BRM has purchased non-QM loan pools from nine sellers. It may onboard additional sellers to the platform that meet its targeted profile of proven, larger originators with a sufficient operating history. Prospective sellers are reviewed via a multistage process and must be approved unanimously by the operational committee, which consists of three teams (pricing and portfolio management, transaction management and operations, and credit and legal) prior to being submitted to the management committee, which comprises senior members of Angelo Gordon's structured credit team, for final approval. In addition to this approval, BRM engages a third-party firm to conduct an operational review of sellers with significant volume. Although BRM has a short history of purchasing non-QM loans, the parent company and the management team have extensive experience purchasing re-performing and non-performing loans, in addition to other real estate assets.

Our overall ranking reflects our qualitative and quantitative review of BRM. Our qualitative review is based on our assessment of three primary focus areas covered during the operational review:

- Management and organization, which includes risk management and financial position;

- Loan purchase and aggregation, which includes property valuation processes; and

- Internal controls, which encompasses operational reviews of originators, pre-purchase data quality, post-purchase quality control, and regulatory compliance.

For our quantitative analysis, we reviewed acquisition volume, loan characteristics, and loan performance history, including delinquencies, early payment defaults (EPDs), and repurchases.

Our AVERAGE qualitative subranking and LIMITED quantitative subranking reflects our assessment of the following strengths and weaknesses.

Strengths:

- The management team's averaged industry experience of over 20 years.

- BRM's relationship with and support from Angelo Gordon, an established investor in credit and real estate with over $43 billion in AUM across a broad range of strategies and experience in reperforming and nonperforming loans.

- Due diligence is conducted on 100% of loans and includes a full review of credit, compliance, property valuation, and fraud.

- The robust internal controls: critical risk decisions require unanimous approval of the operational committee members or subsequent approval by the management committee.

Weaknesses:

- BRM's limited experience in acquiring non-QM loans and limited performance history on its non-QM portfolio. The company started acquiring non-QM loans in December 2017 and has no performance data through a housing cycle downturn.

www.standardandpoors.com February 25, 2021 21

- The post-purchase review process is limited, though this is mitigated via a pre-purchase third-party due diligence review of all loans.

Mortgage originator concentration

In addition to an MOA of BRM (and its application to BRM II for purposes of this transaction), we reviewed historical performance data for loans from originators with exposures above 20% in the GCAT 2021-NQM1 pool. Specifically, loans purchased from Quontic Bank, Arc Home LLC, and HomeXpress Mortgage Corp., which represent approximately 31.9%, 29.8%, and 28.9% of the pool balance, respectively.

Based on the results of this analysis, we concluded that the historical performance of these loans was comparable to the performance of the loans in the BRM and BRM II total portfolio.

We applied a 1.05x to the loss coverage at all rating categories to reflect our view of BRM II's residential mortgage acquisition platform.

Third-Party Due Diligence Review

The third-party due diligence providers-- Clayton Services LLC, Consolidated Analytics LLC, and Digital Risk LLC (which are all on our list of reviewed providers)--performed due diligence on 100% of the pool's loans. Their reviews encompassed credit (underwriting) compliance, property valuations, regulatory compliance, and data quality.

Some loans fell within the scope of the TILA-RESPA Integrated Disclosure (TRID) rule. For these loans, the third-party firms followed the Structured Finance Assn. (SFA) RMBS 3.0 TRID Compliance Review Scope in conducting their final loan reviews (see "Standard & Poor's Comfortable With SFIG Draft Proposal Regarding TRID Due Diligence," published April 25, 2016). In accordance with our criteria, we adjust our loss expectations based on our view of the firms' findings (see Appendix III of "Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later," published Feb. 22, 2018).

Highlights of some of the findings include the following:

Compliance with underwriting guidelines (credit review)

- Most of the due diligence firms' initial findings were resolved by the receipt of missing information in the trailing documents.

- All loans in the pool received a credit risk grade of 'A' or 'B'.

Property valuation review

- All loans received a final property valuation review risk grade of 'A' or 'B'.

Regulatory compliance review

- Most of the due diligence firms' initial findings were resolved by the receipt of missing information in the trailing documents.

www.standardandpoors.com February 25, 2021 22

- Many exceptions resulted from missing documents to complete the ATR/QM or TRID reviews.

All but two loans that were subject to a compliance review received a final compliance review risk grade of 'A' or 'B'. The aforementioned two loans had TRID rules findings, in which case we added $34,000 to our loss severity estimate at all rating categories; therefore, they are exempt from consumer regulations.

Data quality review

- A final tape was provided with updated/corrected data, and no outstanding data issues were noted.

After reviewing the third-party due diligence results, we applied a neutral adjustment of 1.00x to the loss coverage at all rating categories.

Representations And Warranties

Our review of the R&Ws for GCAT 2021-NQM1 focused on whether the representations made by the R&W providers (the originators) were substantially consistent with the set of representations we published as part of our criteria (see Appendix IV of "Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later," published Feb. 22, 2018). We evaluated the strength of these R&Ws and considered whether any breach could have a materially adverse impact for the certificateholders. If the R&Ws in the transaction documents do not address the issues in our published R&W framework, we will determine whether we believe it is appropriate to assess additional credit enhancement. Lastly, we will consider the R&W providers' ability to fulfill their obligations in the event of a breach.

The collateral pool consists of loans from four originators. In this transaction, the originators make the R&Ws with the sponsor backstopping the originator in the event of the originator's insolvency. We reviewed the representations made by the representation providers that were assigned to the trust or restated by the sponsor for the benefit of the certificateholders. The originators, like the sponsor, are unrated entities that may be financially unable to repurchase loans if the need arises. We consider the R&W framework to be weaker than the R&W framework seen in prime jumbo transactions because the review of the breach is not automatic, but is instead, at the option of the controlling holder (initially, the affiliate of the sponsor)--unless a violation of the ATR rules is raised on a loan that suffers a realized loss. In this case, the controlling holder has to review the loan for a breach. However, the R&Ws are in line with other non-QM RMBS transactions.

The R&Ws generally are consistent with our criteria. Knowledge qualifications are limited and do not significantly alter the R&W or remedy. The R&Ws are generally made as of the closing date and, where applicable, the sponsor provides representations during any gap period. The R&W will remain in effect for the transaction's life. However, given that the statute of limitations for R&W claims under New York law is generally six years from the date a representation is made, there could effectively be an expiration date on the R&Ws. The early payment default covenant is in line with what we typically look for in a rated transaction, which generally calls for the R&W provider to repurchase a mortgage loan when the borrower fails to make any of the first three monthly payments due after the origination date unless the delinquency resulted from a servicing issue that has subsequently been, or will be, corrected. In GCAT 2021-NQM1, if a borrower fails to make any of the first three monthly payments (not including any payment subject to a forbearance due to the COVID-19 pandemic) by the last day of the month it is due (or in some cases fails to make the first payment by the last day of the month it is due), measured generally from BRM's purchase

www.standardandpoors.com February 25, 2021 23

date, the applicable originator is required to repurchase the loan. Early payment default related repurchases are not backstopped.

The applicable originator and/or sponsor must appropriately remedy any ensuing R&W breach if it has a materially adverse impact on the loan by curing the breach, or repurchasing the mortgage loan or REO property, as applicable, at the purchase price. The enforcement mechanism for R&W breaches includes provisions for a breach review at the option of the controlling holder--either by an independent reviewer or by the controlling holder itself--for any loan that experiences a realized loss. However, loans with realized losses for which ATR notices have been received are automatically reviewed. If a loan is judicially determined to have a TRID finding, or if the custodian determines a loan has defective or missing documentation, it must be cured or repurchased.

Dispute resolutions are ultimately subject to arbitration proceedings, if necessary, in case the originator and/or the sponsor disputes the controlling holder's determination. The costs of the arbitration will be paid by the controlling holder, subject both to reimbursement for those expenses from the assets of the issuing entity and the annual cap, to the extent the controlling holder wins the arbitration.

The transaction also has a provision in which 25.0% of the certificateholders (the directing certificateholders) can start a review on a loan with a realized loss in case the controlling holder either does not review the loan for a breach or determines that the related originator is not obligated to cure/repurchase the loan. Any costs for that review will be borne by these directing certificateholders and may be recovered from the originator/sponsor if, eventually, the directing certificateholders prevail following any arbitration. However, if the originators do not reimburse the costs incurred by the directing certificateholders, those costs may be reimbursed from the trust.

Although the MOA's result reflects a solid aggregation platform, in our view, parties with potentially limited repurchasing ability are providing R&Ws. Therefore, in conjunction with the weaknesses in the framework as identified above, we applied a 1.10x loss coverage adjustment at all rating levels to compensate for these risks. We believe this adjustment is appropriate in the context of the 100% due diligence performed on the loans and the collateral's relative credit quality.

Cash Flow And Scenario Analysis

We reviewed the transaction structure and performed a cash flow analysis to simulate various rating stress scenarios to determine the preliminary ratings for each class consistent with our criteria, accounting for the available credit enhancement (see tables 11 and 12). We analyzed a variety of scenarios for each rating category, including combinations of the following:

- Front- and back-loaded default timing curves;

- Two-year recovery lag assumptions;

- Fast and slow prepayment assumptions;

- High, low, and forward interest rate curve assumptions; and

- Delinquency assumptions to stress liquidity for potential forbearance.

www.standardandpoors.com February 25, 2021 24

Table 11

Cash Flow Assumptions

Scenario

AAA AA A BBB BB B

Recovery lag (mos.) 24 24 24 24 24 24

Servicer advancing No advancing on loans during the recovery lag period.

Prepayments (%)(i)

Low CPR 1 2 3 4 5 6

High CPR 20 20 20 20 20 20

Scenario 1: Standard delinquency curve for testing triggers Delinquency curve without cash flow stress.

Scenario 2: Delinquencies at 35% for first six months to stress Delinquency curve liquidity and triggers followed by standard delinquency curve to test triggers.

Foreclosure 37.52 32.97 26.56 20.30 14.25 8.36 frequency (%)

Loss severity (%) 47.31 42.31 32.76 27.34 23.16 19.14

Loss coverage (%) 17.75 13.95 8.70 5.55 3.30 1.60

(i)Using a standard prepayment convention. CPR--Conditional prepayment rate. N/A--Not applicable.

Notwithstanding the use of excess interest within the transaction structure, we applied front- and back-loaded rather than bulleted (e.g., a semiannual or an annual lump sum) default timing curves in our analysis. This reflects our view of the potential volatility of cash flows, given the loans are newly originated by a reviewed aggregator, subject to third-party due diligence and include structural considerations, such as pro rata principal allocations among classes A-1, A-2, and A-3, and partial P&I advancing by the servicer.

We applied the foreclosure frequencies, loss severities, and combinations of the stresses noted above in our cash flow runs, and we observed some periodically missed interest due to the liquidity stress associated with no advancing. To pass our applicable rating-specific stresses, the interest carryforward amounts resulting from any missed interest payments on the securities have to be paid in full by the maturity date. All carryforward interest was paid back with interest under the applicable rating-specific stresses in our cash flow projections. The results show that each preliminary rated class in the transaction is enhanced to a degree consistent with the assigned preliminary ratings.

Table 12

Structural Assessment

Percentage point difference Preliminary Initial class Initial credit Loss coverage between credit enhancement and Class rating size (%) enhancement (%) (%) loss coverage

A-1 AAA (sf) 75.25 24.75 17.75 7.00

A-2 AA (sf) 5.95 18.80 13.95 4.85

A-3 A (sf) 11.35 7.45 8.70 (1.25)

M-1 BBB (sf) 3.85 3.60 5.55 (1.95)

B-1 BB (sf) 2.00 1.60 3.30 (1.70)

www.standardandpoors.com February 25, 2021 25

Table 12

Structural Assessment (cont.)

Percentage point difference Preliminary Initial class Initial credit Loss coverage between credit enhancement and Class rating size (%) enhancement (%) (%) loss coverage

B-2 B (sf) 1.25 0.35 1.60 (1.25)

B-3 NR 0.35 0.00 N/A N/A

NR--Not rated. N/A--Not applicable.

Servicer stop advance stresses

Although the transaction documents provide for up to three months of P&I advance obligation, we assumed that no P&I advances were being made in our cash flow projections. This assumption results in no projected monthly cash flows on defaulted loans that have not yet been liquidated (we assume a 24-month lag between default and liquidation). Our cash flow projections take into account these liquidity stresses and the transaction's ability to make monthly interest payments and, if necessary, interest carryforward amounts by the final maturity date on the preliminary rated classes. We also modeled a delinquency curve, based on our criteria, for the purpose of testing the delinquency trigger.

To address the potential liquidity stress to cash flows due to loans entering forbearance in light of the current COVID-19 pandemic crisis for which the P&I advancing party is not obligated to advance monthly P&I payments, we applied an additional delinquency stress scenario. We assumed 35.00% of the closing pool balance to be delinquent for the first six months with any P&I payments related to this delinquent portion coming back to the transaction after all defaults have been passed through to the transaction (approximately 144 months).

WAC deterioration stress

The transaction structure allows excess spread to provide some of the credit enhancement. We applied a WAC deterioration stress that steps up linearly from zero basis points (bps) to 82 bps over 10 years. It remains at that level going forward to address the potential for the pool's WAC to decline as higher coupon loans prepay or default and thus stress the excess spread.

Interest stresses

All of the rated certificates have coupons subject to the net WAC-rate cap, as is the case for most post-2009 transactions that we have rated. If the net WAC rate decreases below the cap, the interest due to the certificates will decrease by a similar amount.

In this transaction, extraordinary trust expense payments reduce the net WAC rate, which effectively allocates the extraordinary trust expenses pro rata across all senior and subordinate certificateholders by reducing their interest payments by the amount of the extraordinary trust expenses paid (subject to the annual cap). Although the extraordinary trust expenses are passed through as reduced contractual interest due to certificateholders, we ran these expenses from period 13 to 60 (four years) at a certain percentage of the capped amounts as specified by our criteria to test any impact on the securities due to their dependence on excess spread as a form of credit enhancement and the presence of certain structural features, such as limited P&I

www.standardandpoors.com February 25, 2021 26

advancing. We also took this approach because interest payments on the securities are deferrable.

Imputed Promises Analysis

We impute the interest owed to the security holders when rating U.S. RMBS transactions where credit-related events can reduce interest owed to the tranches across the capital structure rather than an allocation of that credit-related loss to the available credit support, based on our loan modification guidance, "Guidance: Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later," published Dec. 8 2020. WAC deterioration that occurs because of defaults, repurchases, or prepayments is not considered credit-related; therefore, it is not considered as part of this analysis.

This transaction provides for credit-related loan modifications and extraordinary trust expenses to reduce the net WAC, at which the transaction's bond coupons are capped. Therefore, we applied the approach outlined in the criteria to assess the maximum potential rating (MPR) that could apply, based on our projected interest reduction amount (PIRA). Because this is a new issue transaction, we did not account for any cumulative interest reduction amount.

Consistent with our criteria, we assumed that 50.0% of the loans projected to default would be modified. We also assumed that 75.00% of the projected modifications are interest rate modifications, with an interest rate reduction of 2.00%. When added to the extraordinary trust expenses, this resulted in a maximum PIRA on the preliminary rated certificates that is below the 4.50% threshold. We stressed extraordinary trust expenses by the relevant extraordinary expense application factor over four years from payment periods 13 to 60. Based on the results of our analysis, there was no impact on the securities' MPR.

Historically, we have observed that extraordinary trust expenses have been both minimal when they occur and extremely limited in pre-2009 RMBS transactions. We continue to expect their actual occurrence in post-2009 transactions to be rare.

Operational Risk Assessment

Our criteria "Global Framework For Assessing Operational Risk In Structured Finance Transactions," published Oct. 9, 2014, present our methodology and assumptions for assessing certain operational risks (severity, portability, and disruption) associated with asset types and key transaction parties (KTPs) that provide an essential service to a structured finance issuer. As outlined in the criteria, we cap the ratings on a transaction if we believe operational risk could lead to credit instability and affect the ratings.

As provided in the operational risk criteria, for severity risk and portability risk, there are three possible rankings: high, moderate, or low. For disruption risk, there are four possible rankings: very high, high, moderate, or low. The rankings for each of the risks determine the MPR that can be assigned to a structured finance security for a given KTP before giving consideration to any provisions for a backup KTP, such as a master servicer.

According to our criteria, we rank severity and portability risk for nonprime residential mortgage collateral as moderate and low, respectively. For GCAT 2021-NQM1, the servicer, Shellpoint, is the KTP. We assess the disruption risk for Shellpoint as low. Given these risk assessments, our criteria does not cap the ratings on the transaction.

www.standardandpoors.com February 25, 2021 27

Related Criteria

- Criteria | Structured Finance | General: Global Framework For Payment Structure And Cash Flow Analysis Of Structured Finance Securities, Dec. 22, 2020

- Criteria | Structured Finance | General: Methodology To Derive Stressed Interest Rates In Structured Finance, Oct. 18, 2019

- Criteria | Structured Finance | Legal: U.S. Structured Finance Asset Isolation And Special-Purpose Entity Criteria, May 15, 2019

- Criteria | Structured Finance | General: Counterparty Risk Framework: Methodology And Assumptions, March 8, 2019

- Criteria | Structured Finance | General: Incorporating Sovereign Risk In Rating Structured Finance Securities: Methodology And Assumptions, Jan. 30, 2019

- Criteria | Structured Finance | RMBS: Assumptions Supplement For Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later, Feb. 22, 2018

- Criteria | Structured Finance | RMBS: Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later, Feb. 22, 2018

- Criteria | Structured Finance | RMBS: U.S. Residential Mortgage Operational Assessment Ranking Criteria, Feb. 22, 2018

- General Criteria: Methodology For Linking Long-Term And Short-Term Ratings, April 7, 2017

- Criteria | Structured Finance | General: Global Framework For Assessing Operational Risk In Structured Finance Transactions, Oct. 9, 2014

- General Criteria: Global Investment Criteria For Temporary Investments In Transaction Accounts, May 31, 2012

- General Criteria: Principles Of Credit Ratings, Feb. 16, 2011

- Criteria | Structured Finance | General: Methodology For Servicer Risk Assessment, May 28, 2009

Related Research

- Select Servicer List, Feb. 3, 2021

- U.S. Residential Mortgage And Housing Outlook: Positive Momentum Carries Into 2021, Jan. 22, 2021

- S&P Global Ratings Definitions, Jan. 5, 2021

- Servicer Evaluation: Nationstar Mortgage LLC, Dec. 22, 2020

- Economic Research: Staying Home For The Holidays, Dec. 2, 2020

- S&P Global Ratings Publishes List Of Third-Party Due Diligence Firms Reviewed For U.S. RMBS As Of Nov. 5, 2020, Nov. 5, 2020

- Can COVID-19 Cause A Cash Crunch For Certain U.S. RMBS?, Aug. 21, 2020

- Non-Qualified Mortgage Loans Summertime Blues Continue Despite Improved July

www.standardandpoors.com February 25, 2021 28

Delinquencies, July 29, 2020

- Non-QM RMBS And COVID-19: Locking Down States' Exposure, June 1, 2020

- Guidance: Methodology And Assumptions For Rating U.S. RMBS Issued 2009 And Later, April 17, 2020

- Servicer Evaluation: Shellpoint Mortgage Servicing, April 1, 2020

- Assessing The Credit Effects Of COVID-19 On U.S. RMBS, March 20, 2020

- U.S. Residential Mortgage Input File Format For LEVELS, March 6, 2020

- Credit Rating Model: LEVELS Model For U.S. Residential Mortgage Loans, Aug. 5, 2019

- Key Factors For Assessing U.S. Non-Qualified Mortgage Bank Statement Loans, April 10, 2019

- Credit Rating Model: Intex RMBS Cash Flow Model, April 7, 2017

- Global Structured Finance Scenario and Sensitivity Analysis 2016: The Effects of The Top Five Macroeconomic Factors, Dec. 16, 2016

- Standard & Poor's Comfortable With SFIG Draft Proposal Regarding TRID Due Diligence, April 25, 2016

- Older RMBS Transactions Face Increased Tail Risk As Their Pools Shrink, Aug. 9, 2012

www.standardandpoors.com February 25, 2021 29

No content (including ratings, credit-related analyses and data, valuations, model, software or other application or output therefrom) or any part thereof (Content) may be modified, reverse engineered, reproduced or distributed in any form by any means, or stored in a database or retrieval system, without the prior written permission of Standard & Poor's Financial Services LLC or its affiliates (collectively, S&P). The Content shall not be used for any unlawful or unauthorized purposes. S&P and any third-party providers, as well as their directors, officers, shareholders, employees or agents (collectively S&P Parties) do not guarantee the accuracy, completeness, timeliness or availability of the Content. S&P Parties are not responsible for any errors or omissions (negligent or otherwise), regardless of the cause, for the results obtained from the use of the Content, or for the security or maintenance of any data input by the user. The Content is provided on an “as is” basis. S&P PARTIES DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE OR USE, FREEDOM FROM BUGS, SOFTWARE ERRORS OR DEFECTS, THAT THE CONTENT'S FUNCTIONING WILL BE UNINTERRUPTED OR THAT THE CONTENT WILL OPERATE WITH ANY SOFTWARE OR HARDWARE CONFIGURATION. In no event shall S&P Parties be liable to any party for any direct, indirect, incidental, exemplary, compensatory, punitive, special or consequential damages, costs, expenses, legal fees, or losses (including, without limitation, lost income or lost profits and opportunity costs or losses caused by negligence) in connection with any use of the Content even if advised of the possibility of such damages.

Credit-related and other analyses, including ratings, and statements in the Content are statements of opinion as of the date they are expressed and not statements of fact. S&P's opinions, analyses and rating acknowledgment decisions (described below) are not recommendations to purchase, hold, or sell any securities or to make any investment decisions, and do not address the suitability of any security. S&P assumes no obligation to update the Content following publication in any form or format. The Content should not be relied on and is not a substitute for the skill, judgment and experience of the user, its management, employees, advisors and/or clients when making investment and other business decisions. S&P does not act as a fiduciary or an investment advisor except where registered as such. While S&P has obtained information from sources it believes to be reliable, S&P does not perform an audit and undertakes no duty of due diligence or independent verification of any information it receives. Rating-related publications may be published for a variety of reasons that are not necessarily dependent on action by rating committees, including, but not limited to, the publication of a periodic update on a credit rating and related analyses.

To the extent that regulatory authorities allow a rating agency to acknowledge in one jurisdiction a rating issued in another jurisdiction for certain regulatory purposes, S&P reserves the right to assign, withdraw or suspend such acknowledgment at any time and in its sole discretion. S&P Parties disclaim any duty whatsoever arising out of the assignment, withdrawal or suspension of an acknowledgment as well as any liability for any damage alleged to have been suffered on account thereof.

S&P keeps certain activities of its business units separate from each other in order to preserve the independence and objectivity of their respective activities. As a result, certain business units of S&P may have information that is not available to other S&P business units. S&P has established policies and procedures to maintain the confidentiality of certain non-public information received in connection with each analytical process.

S&P may receive compensation for its ratings and certain credit-related analyses, normally from issuers or underwriters of securities or from obligors. S&P reserves the right to disseminate its opinions and analyses. S&P's public ratings and analyses are made available on its Web sites, www.standardandpoors.com (free of charge), and www.ratingsdirect.com and www.globalcreditportal.com (subscription), and may be distributed through other means, including via S&P publications and third-party redistributors. Additional information about our ratings fees is available at www.standardandpoors.com/usratingsfees.

Standard & Poor’s | Research | February 25, 2021 30 2601268