CurveGlobal Interest Rate Derivatives SONIA Futures: New Opportunities Andy Shaw and Dr. Mark Tomsett, Links Risk Contents

1 Introduction 2 2 What is SONIA? 2 2.1 The End of ? 2 2.2 How is it calculated? 2 2.3 How is it used? 3 2.4 What is an OIS? 3 2.5 What influences OIS rates? 5 3 SONIA Futures 6 3.1 Contract Specifications 6 3.2 Basis Point Value 7 3.3 Futures vs OIS 7 3.3.1 Convexity: An Introduction 8 3.3.2 Convexity: In Detail 9 3.3.3 Monetising Convexity 9 3.4 Spread Trading Strategies 10 3.5 Managing Portfolio Delta Risk 10 4 Miscellaneous Items 13 4.1 Liquidity 13 4.2 Efficiency 13 5 Summary 14 6 Definitions 15 References 16 Disclaimer 17

curveglobal.com | linksriskadvisory.com 1 1 Introduction

This note takes a look at the creation of interest rate exposure using exchange traded derivatives, in particular the SONIA-indexed . Given the regulatory driven changes that are starting to facilitate switches away from benchmarks like LIBOR, and with benchmarks like SONIA for GBP denominated derivatives gaining in popularity, there is increased demand for new products referencing this key index and information explaining their behaviour.

2 What is SONIA?

2.1 The End of LIBOR?

From the end of 2021 the Financial Conduct Authority (FCA) will no longer compel panel to submit London Interbank Offered Rate (LIBOR) estimates. Some evidence¹ suggests that the underlying market for unsecured interbank overnight lending, which LIBOR purports to measure, is no longer sufficiently liquid. In addition, the EU Benchmark Regulation introduced in 2018, which necessitates benchmarks to represent the underlying market being measured, further requires that input data based on discretion and expert judgment be subject to enhanced procedures and controls. As some LIBOR panel banks provide input based on expert judgment alone, panel banks may consider the legal and regulatory risks too great to continue submissions once no longer compelled to do so. In April 2017 the of England (BoE) Risk Free Working Group selected the Sterling Overnight Index Average (SONIA) as its proposed alternative benchmark.

2.2 How is it calculated?

SONIA, originally introduced in 1997, has been administered by the BoE since 2016, and in a modified form since April 2018. It is²:

A measure of the rate at which interest is paid on sterling short-term wholesale funds in circumstances where credit, liquidity and other risks are minimal.

It is the trimmed, volume-weighted mean of interest rates paid on eligible sterling denominated deposit transactions. Eligible transactions have to meet the following conditions: to be reported within a specific time-frame, unsecured and of one business day maturity, executed within 00:00 and 18:00 UK time and have a minimum notional value of £25 million.

1See [4] and [7] 2See [1]

curveglobal.com | linksriskadvisory.com 2 2.3 How is it used?

Overnight benchmarks, like SONIA, are generally referred to as overnight indexed (OIS) rates, because they are used primarily for the settlement of interest rate swaps. They are also important in determining the amount of cash that is paid on collateral posted to secure over-the-counter (OTC) liabilities between counterparties. For example, a central counterparty (CCP) will transfer money between two members to reflect changes in mark-to-market (MtM) on a daily basis. If a transaction moves in favour of one member they will receive in their collateral account a cash amount, known as variation margin (VM), equal to the change in value. This amount is debited from the collateral account of the other member. The VM recipient is required to pay what is known as Price Alignment Interest (PAI) to the VM payer. PAI is usually set using an overnight index.

2.4 What is an OIS?

An OIS is an interest rate-sensitive derivative, very similar to a standard (IRS) but with one important difference: the determination of the floating rate. An IRS will have a single (IBOR-based) fixing per floating rate period that determines the relevant cashflow. This means that the payment can be brought forward because, once the IBOR rate index is published at the start of the period, the cashflow is known and can be paid at any point. The same does not apply to an OIS, where all the daily fixings over the floating rate period are required before the final payment is known. OIS tend to have annual fixed payments and annual floating frequencies³. That means something in the order of 252 separate fixings. This has implications for how the final leg is calculated and the relationship between the fixing date, the publication date, the effective date and the maturity date. The floating leg payment for the single period of an IBOR- indexed IRS is dependent on the information shown in Figure 1. The fixing date and publication date are the same day. The start date (for the calculation of the accrual) will either be the same day or one or two days later, depending on the currency convention. Note that the settlement takes place on the start date.

Figure 1: IBOR Floating Leg Schedule

3Or single payments at maturity for those with a tenor of less than 1 year.

curveglobal.com | linksriskadvisory.com 3 For a SONIA-indexed swap, for a given fixing date, the publication date is the next good business date at 09:00. The effective date is the fixing date and the maturity date is the next good business date. For SONIA-indexed swaps there is no lag between the maturity date and the settlement date.

The final payment is generated by compounding the overnight interest rate over the whole floating period. Equation 1 shows the calculation of the fixed leg (left-hand side) and the floating leg (right-hand side) of a single payment OIS.

Where:

TE = start date of the swap

TM = maturity date of the swap

TS = settlement date of the swap

ROIS = OIS rate

D = day count basis

ri = the on day i

ni = the number of days over which ri compounds

DFTS = discount factor to date TS

∏ = the product of a sequence of numbers

ni will typically be equivalent to 1 day but over a weekend, or holiday, may be longer.

curveglobal.com | linksriskadvisory.com 4 Figure 2: OIS Floating Leg Schedule

The term in the square brackets of equation 1 is the compounded SONIA rate. The compound rate ( r˜ ) equivalent to the individual SONIA fixings can be calculated using equation 2.

For example, assume that a floating leg is based on three SONIA fixings published on Friday, Monday and Tuesday. The fixings are 1.5%, 1.55% and 1.6% respectively. The compound rate between Friday and Wednesday is calculated as:

Notice how, due to the weekend, the first fixing applies over three days rather than just one. Weekends and official holidays, for example Easter, can increase the importance of fixings over these periods, particularly for shorter dated trades.

4 The payment of this leg takes place on the last publication date of the index , so TM = TS. The timeline of a single fixing of a generic OIS floating leg is shown in Figure 2.

2.5 What influences OIS rates?

As indicated in section 2.2, unlike LIBOR, SONIA is calculated from actual transactions. The qualifying transactions are only those made by ‘recognised’ banks: those who have been approved by the BoE to use its lending and deposit facilities. It is the rate that is paid on overnight deposits to these banks that is a key determinant of other short-term interest rates. Theoretically, being backed by the BoE, it should represent a virtually risk-free rate and act as a floor to all other rates. However, there are some institutions who would like to use the BoE deposit facilities but are unable due to being ‘unrecognised’. Instead they deposit funds with recognised banks who in turn deposit with the . In order to profit from this situation the recognised bank will offer the lender a rate below the BoE deposit rate. When the supply of liquidity from unrecognised banks falls, the rate offered will rise. Therefore, by a quirk of the UK market structure, as the supply of funds fluctuates, SONIA can trade above and below the BoE deposit rate5.

4In other currencies this payment may take place with a lag, which introduces a potential convexity issue. This has been shown to be negligible. See [6]. 5See [3] for a more detailed discussion.

curveglobal.com | linksriskadvisory.com 5 3 SONIA Futures

3.1 Contract Specifications

On 30th April 2018 CurveGlobal6 (CG) launched its Three-month SONIA Futures contract. The contract specification has the following key features7:

• SONIA indexed.

• Contract Notional £500,000, Fixed Tick Size 0.005, Tick Value £6.25.

• Accrual Period - The contract month’s International (IMM8) date to the day before the next IMM date.

»» For example, the June contract accrual will start on the June IMM date and end on the September IMM date.

• The contract will expire/mature on the third Wednesday of the IMM month.

• The EDSP9 of the futures contract is 100 – R, where R is based on the compounded SONIA rate over the Accrual Period.

One of the attractions of the Three-month SONIA futures contract is that it has been deliberately designed in a way that replicates the floating leg construction of the SONIA OIS.

6CurveGlobal is an interest rate derivatives venture between the London Stock Exchange Group, the Cboe and a number of the major dealer banks. See the CurveGlobal website for details. 7See the full product description on the CurveGlobal website. 8IMM dates are the four quarterly dates: March, June, September and December. 9Exchange Delivery Settlement Price.

curveglobal.com | linksriskadvisory.com 6 3.2 Basis Point Value

The change in the value of a futures contract is closely tied to the concept of Basis Point Value (BPV), which in turn is a fundamental component of managing risk. The BPV for a 3-month (3M) futures contract is calculated using the following formula:

BPV = Face Value × 3M accrual × 1 basis point (3)

By convention the 3M accrual is assumed to be 0.25. Therefore, the BPV for this contract is: 1 £500,000 × 0.25 × = £12.50 10000 Suppose the Three-month SONIA futures contract is trading at 99.15. This implies a 3M compounded SONIA rate of 100 - 99.15 = 0.85%. If the price rises to 99.35 this implies a new rate of 100 - 99.35 = 0.65%. In other words the implied rate has fallen 0.2%, or 20 basis points. What does this mean in terms of the change in the value of the contract?

For each contract the total value of this change is equal to:

No of contracts × BPV × no of basis points (4)

In this example, for a single contract, this is equal to:

1 × £12.50 × 20 = £250

Note how someone who has sold a SONIA futures contract (i.e. ‘short’) profits from a fall in the contract price/rise in rates while someone who has bought the same contract (i.e. ‘long’) profits from a rise on the contract price/fall in rates.

3.3 Futures vs OIS

The demise of LIBOR as noted in section 2.1, together with the increasing collateralisation of interest rate derivatives and our better understanding of the associated credit risks, has provided an impetus to the development of a whole new suite of interest rate products. The Three-month SONIA future allows both direct interest rate exposure and provides a hedge for exposure created using other interest rate products.

Consider the following simple example of an investor who is short a 3M SONIA-indexed OIS, i.e. lending at the fixed rate. The natural hedge of this position is to sell a matching Three-month SONIA futures contract. The payoffs of the two positions are given below:

Where F is the OIS fixed rate/the rate implied by the futures purchase price, and L is the maturity date OIS fixing/the EDSP implied rate. Table 1 shows some sample details. Note that the notional of the OIS has to be slightly higher than the total notional value of the futures so that the two deltas (sensitivity to interest rate changes expressed in GBP) are equivalent. This is determined by solving equation 7 for

NOIS, the OIS notional.

Where:

NOIS = OIS Notional

∆OIS = the sensitivity of the OIS to a 1 basis point move in the fixed rate

Φ = the number of futures contracts

NF = the face value of each futures contract

BPV = the Basis Point Value

curveglobal.com | linksriskadvisory.com 7 The two trades are assumed to have been executed at the same rate.

Table 1: Trade Details

Contracts Face Value Total Notional Price Rate

Future 1,000 500,000 500,000,000 99.15 0.85%

OIS 501,050,000 0.85%

Table 2 shows the payoffs for each trade for a variety of different fixing levels. The future is a good hedge for the swap, particularly for small changes to the fixing rate. The table also shows an example of the ‘convexity bias’ between the two positions. When rates fall the OIS gains slightly more than the futures lose. When rates rise the OIS loses slightly less than the futures gain. This is a function of the discounting that is applied to the OIS payoff but not the future. Let’s consider this in more detail.

Table 2: Trade Payoffs at Different Fixings

Rates 0.55% 0.65% 0.75% 0.85% 0.95% 1.05% 1.15%

Future -375,000 -250,000 -125,000 0 125,000 250,000 375,000

OIS 375,272 250,119 125,028 0 -124,966 -249,869 -374,710

Net 272 119 28 0 34 131 290

3.3.1 Convexity: An Introduction

Convexity is a term used in numerous different contexts and situations in finance, sometimes correctly, sometimes not. Roughly speaking, convexity is normally used to describe the non-linearity of certain contracts in terms of changes of value versus unit changes in rates up or down. In this case, an OIS swap demonstrates convexity in its price behaviour as it is a discounted instrument. Whereas a future does not exhibit this same behaviour as it is not a discounted instrument. Convexity adjustments are required when trying to compare futures implied rates to implied rates from discounted instruments to establish if the markets are aligned. Convexity adjustments can be used to describe a broad range of effects beyond this highlighted example in other instruments. In other cases, and especially when considering options in a multi-curve world, convexity adjustments can get very esoteric in nature. However, for futures versus swaps, IBOR or OIS, the convexity issue currently is very small due to the low rate environment and generally vanilla in nature.

curveglobal.com | linksriskadvisory.com 8 3.3.2 Convexity: In Detail

Convexity is the term widely used to describe the gamma risk associated with either an interest rate-sensitive instrument, or perhaps an interest rate portfolio. Gamma risk is the rate of change of delta; a measure of how much delta changes as interest rates change. If delta is the first derivative of the present value (PV) with respect to movements in the underlying rates, gamma is the second derivative.

In the example above, an investor can either buy the OIS/pay fixed or sell the OIS/receive fixed. Buying an OIS incurs negative gamma risk and selling an OIS incurs positive gamma. Why is this10?

An investor who has sold the OIS will make an MtM loss as rates rise and an MtM gain as rates fall, due to the relative difference between the fixed and floating rates on the swap. There is only a single cashflow (at least for a single period OIS) and the PV of this MtM change is a function of the discount factor, which is also impacted by the change in rates. If rates rise the discount factor falls, so the MtM loss is not quite so bad in PV terms because it is discounted more heavily. As rates fall, the discount factor increases so the MtM gain is discounted less heavily. So the seller of the swap benefits whether rates rise or fall, hence the positive gamma. The converse would be true for a buyer of the swap. A longer dated instrument, or one that is forward starting, would typically show higher levels of gamma risk. As relatively short dated instruments, compared to LIBOR-indexed IRS, OIS tend to exhibit relatively low levels of gamma risk.

In the example shown the OIS has gamma due to the fact that its delta changes as the market moves. The futures position does not have gamma; its sensitivity to rate changes is constant (the BPV). This portfolio of swaps and futures would have to be re-hedged as the total delta changes. The short OIS position has positive gamma which generates positive P&L in the overall portfolio. Therefore, there is an economic advantage to lending via the OIS rate (by selling) and borrowing via the future (by selling). Selling pressure on the futures contract means that futures rates are higher than those implied by the equivalent OIS. This differential is known as the Futures Convexity Adjustment (FCxA). It is the upfront value of the total amount of P&L that a portfolio would be expected to accrue until both the OIS and the hedging future expire simultaneously. In theory an adjustment should be made to reflect this possible arbitrage. However, calculating FCxAs can be a technical challenge11. In addition, the theoretical and actual values of the FCxAs may not necessarily be aligned due to relative changes in the supply and demand of the respective instruments.

3.3.3 Monetising Convexity

The positive gamma shown in this example can be monetised through the CCP margin process. The MtM changes of the future and the swap result in VM flows. The VM flows on the future are treated as final settlement. Therefore, should your counterparty default you will receive no further payment. In contrast, VM flows on the swap are regarded as the present value of future cashflows that remain unsettled until the coupon date. In the event of a default the portfolio of trades owned by the defaulting party are auctioned, such that eventually a new party takes ownership, and the contract continues unchanged until maturity. Therefore, VM payments are regarded as a margin asset that you would not expect to lose in a default, and why PAI is applied.

The key feature of this example is that the short futures position earns money in a rising rate market and loses money in a falling rates market. The VM can then be either reinvested at a higher rate or funded at the new lower rate. In markets where both the futures and the swap position are centrally cleared it becomes less clear whether the convexity trade makes money once margin funding costs and PAI (which may not be the same) are taken into account. In the current low rate environment, the cost of funding margin is often higher than the rates that are paid by the CCP. This is particularly true if the two products are not cross-margined. This may be because the swap and the future are not both cleared by the same CCP, or the CCP doesn’t allow the two to be comingled for margining purposes. This is discussed further in Section 4.2.

10This is a classical description of gamma. For a more detailed explanation see [3]. 11See for example [8].

curveglobal.com | linksriskadvisory.com 9 3.4 Spread Trading Strategies

As well as taking outright interest rate risk and trading convexity bias there are myriad strategies that can be exploited using SONIA futures. One of the most basic, that forms the basis for many other strategies, is the spread trade. Buying and selling futures contracts at different parts of the yield curve, even across yield curves (e.g. SONIA vs 3M Libor) allows an investor to express their view of changes in interest rates. In general, this takes the form of steepeners (where the difference between short term and long term rates widens) and flatteners (where the difference between short and long term rates narrows).

Suppose in February the June Three-month SONIA future is trading at 99.5 and the September future is trading at 99.4. The implied spread is 0.1%. To profit from a steepening curve the investor buys the June future (anticipating a rate fall/price rise) and sells the September future (anticipating a rate rise/price fall). To profit from a flattening curve the reverse trade is executed. This is known as a calendar trade. Buying the nearest to expire contract and selling a contract with a later expiry is known as a /long spread trade. The reverse is known as a /short spread trade.

3.5 Managing Portfolio Delta Risk

As well as being able to hedge single swaps, SONIA futures can also be used to hedge swap portfolios. The standard approach for delta hedging a swap portfolio is to match the delta of the swap portfolio against a weighted sum of hedging instruments. Let y represent the interest rate portfolio to be hedged by a series of M interest rate sensitive instruments, each individually represented as Xi, and each with a weight Wi. Let ∏ represent the combined/hedged portfolio such that:

Where ∑ is the sum of a sequence of numbers. Let rk be the interest rates to which both the swap portfolio and the hedging instruments are sensitive. Then, if the swap portfolio is delta hedged:

Where:

This is typically written in a different form:

Where:

J is known as the Jacobian matrix and JT is its transpose. The weights are therefore found by rearranging equation 10 to give:

The following simple example illustrates this process. Given a portfolio of three off-the-run OIS, of varying maturity (up to 2 years), payment frequencies12, direction and notional, a cashflow table was created. This is shown in table 3.

12Even though they trade with annual cashflows or a single settlement at maturity if the tenor is less than one year.

curveglobal.com | linksriskadvisory.com 10 Table 3: Portfolio Cashflows Table 4: Portfolio Sensitivity and Futures Hedges

Date Cashflow (£) Futures PV01 Hedges

16-Jun-19 -9,972.60 1 239.11 19.1

14-Feb-19 52,931.51 2 -2.32 -0.2

03-Mar-19 -81,326.03 3 3.52 0.3

14-May-19 48,532.19 4 244.97 19.6

14-Aug-19 45,144.72 5 245.19 19.6

03-Sep-19 -78,954.81 6 243.58 19.5

14-Nov-19 36,377.75 7 72.50 5.8

16-Dec-19 -7,451.31 8 -206.66 -16.5

14-Feb-20 25,537.57

03-Mar-20 -60,303.14

14-May-20 14,963.48

14-Aug-20 2,780.50

03-Sep-20 -38,656.77

14-Nov-20 -14,431.20

03-Mar-21 -2,602.20

curveglobal.com | linksriskadvisory.com 11 By valuing a portfolio using a yield curve built from the hedging instruments (in this case SONIA futures) it is possible to calculate the sensitivity of the portfolio to changes in the values of those instruments as well as the sensitivities of the instruments themselves. The sensitivity of the futures is simply the BPV. Table 4 shows the impact on the PV of the portfolio of increasing the price of each future by 1 basis point. With this information it is possible, by exploiting equation 11, to calculate the number of futures contracts (and whether long or short) that are required to hedge these movements. This is shown in table 4.

The effectiveness of using a strip of futures in this way is shown in Figure 3. In this example the combined portfolio benefits from convexity, or positive gamma. The hedge portfolio is only designed to protect against delta risk. In reality the portfolio delta will change at different levels of interest rate and it also fails to account for the fact that changes in rates are rarely independent.

It should also be acknowledged that SONIA-based yield curves do not generally include futures instruments currently. The curve set of liquid instruments is OIS out to 2 years and then basis swaps between the different rate tenors.

Figure 3: Hedge Effectiveness

curveglobal.com | linksriskadvisory.com 12 4 Miscellaneous Items

4.1 Liquidity

The discussion up to this point has made an implicit assumption that there is enough liquidity in this product to make it a viable trading . Since its launch in 2018, the CG Three-month SONIA Futures trading volume has grown to the point where over £43 billion notional has been traded with approximately £9 billion notional outstanding13. This represents about 50% of the market share14. At the moment the preferred method of price discovery at the short-end is the OIS which forms a fundamental tool in yield curve construction. As liquidity grows further this may change but will depend on whether it allows the kind of granularity around key rate change events (e.g. MPC15 meeting dates) that is needed.

4.2 Margin Efficiency

The CG Three-month SONIA Futures trade on CurveGlobal Markets, the of London Stock Exchange plc, and are cleared by LCH. Standalone futures contracts are margined using a 2 day Margin Period of Risk (“MPOR”), whereas OTC products use a 5 day MPOR (or 7 day for clients). A higher MPOR implies a longer period of risk neutralisation and hence a larger Initial Margin amount. If futures are included in an exclusive standalone, futures-only account then they will be margined at the lower MPOR amount; whereas futures in an inclusive, swaps-and-futures account will have the advantage of netting of risks with OTC, but the overall account will have a higher MPOR.

It might therefore seem at first sight that there is a decision to make by LCH members to choose whether to margin their CG positions standalone, or as part of a wider portfolio. Clearly the inputs to that decision are complex and would change as the composition of the portfolio changes, or as the inputs to the IM model changes (The initial margin required by LCH is calculated using an historic simulation model16 for both futures and OTC.) Fortunately for members and clients at LCH they do not need to make this decision. LCH’s Spider17 automatically allocates the required number of futures between the standalone futures account and the combined ETD/OTC account to minimise total margin. Users of CG Three-month SONIA contracts can therefore trade in the most margin-efficient way at LCH; LCH’s Spider will determine the optimum composition of futures in a standalone (low MPOR, no netting) or combined (higher MPOR and netting) account.

Anecdotal evidence suggests that in many instances, when SONIA Futures are included in portfolios, the margin calculation is static after the inclusion of the futures representing an opportunity to save the margin that had previously been posted against the futures in isolation. Although these savings are portfolio specific, it shows there are potential benefits to be gained by entities if risk exposure is cleared thoughtfully.

13See [2]. 14Similar products trade on ICE and the CME. 15The BoE Monetary Policy Committee. 16See the LCH website for details. 17See LCH Spider on the LCH website for details.

curveglobal.com | linksriskadvisory.com 13 5 Summary

The current interest rate market is in a state of change prompted by the move away from the long-established LIBOR benchmarks. There is intense debate over how the fall-back will be managed17. We have already seen a significant change in the way that interest rate curves are constructed, with a dramatic increase in the number of curves required to price an interest rate portfolio. As different jurisdictions settle on their preferred IBOR-replacement we would expect markets to start to develop a coherent suite of yield curve products. In the UK, where the choice of SONIA has already been made, with the creation of instruments like SONIA futures, this conversion process is well underway.

17See for example [5].

curveglobal.com | linksriskadvisory.com 14 6 Definitions

This section provides a list of the key acronyms used in the document:

3M Three Month(s) BPV Basis Point Value BoE Cboe Cboe Global Markets CG CurveGlobal CCP Central Counterparty EDSP Exchange Delivery Settlement Price EU European Union FCxA Futures Convexity Adjustment GBP British Pound IBOR Interbank Offered Rate IMM International Monetary Market IRS Interest Rate Swap LCH London Clearing House LIBOR London Interbank Offered Rate MPC Bank of England Monetary Policy Committee MPOR Margin Period of Risk MtM Mark-to-Market OIS OTC Over-the-Counter P&L Profit and Loss PAI Price Alignment Interest PV Present Value SONIA Sterling Overnight Index Average VM Variation Margin

curveglobal.com | linksriskadvisory.com 15 References

[1] Bank of England. SONIA: Key features and policies. Tech. rep. Bank of England, 2018.

[2] Bank of England Risk Free Working Group. November 2018 Newsletter, 2018. URL: https://www.bankofengland.co.uk/-/media/boe/files/markets/benchmarks/newsletter-november-2018.

[3] Darbyshire, JHM. Pricing and Trading Interest Rate Derivatives: A Practical Guide to Swaps. Aitch & Dee, 2016.

[4] FCA. The future of LIBOR, 2017. URL: https://www.fca.org.uk/news/speeches/the-future-of-libor.

[5] Henrard, Mark. “A quant perspective on IBOR fallback proposals”. In: muRisQ Advisory Working Paper (2018).

[6] Henrard, Mark. “Overnight Indexed Swaps And Floored Compounded Instrument In HJM One-Factor Model”. In: Working Paper (2004).

[7] Risk Magazine. Beyond Libor: Special Report 2018, 2018.

[8] Ron, Uri. “A practical guide to swap curve construction”. In: Bank of Canada Working Paper (2000).

curveglobal.com | linksriskadvisory.com 16 Disclaimer

The “SONIA” mark is used under licence from the Bank of England (the benchmark administrator of SONIA), and the use of such mark does not imply or express any approval or endorsement by the Bank of England. “Bank of England” and “SONIA” are registered trademarks of the Bank of England. By making this communication Curve Global Limited and Links Risk Ltd (the “Parties”) do not intend to invite or induce you to engage in any investment activity for the purposes of the UK regulatory regime (other than, where relevant, in respect of the London Stock Exchange plc’s exempt activities of operating UK-regulated investment exchanges and of LCH Limited providing clearing services in the UK). The materials presented by the Parties to illustrate its products or services (the Materials) are for information purposes only and do not constitute, nor purport to effect, an offering of, or solicitation of offers to purchase or subscribe for, any securities, the provision of investment or trading advice or an intermediation in activities pertaining to the marketing of securities, futures contracts, options on futures, swaps or other commodities. No person should act on the basis of the Materials.

The Materials are being given on the basis that each person in the United Kingdom to whom it is issued is reasonably believed to be such a person as is described in Article 19 (“Investment professionals”) or Article 49 (“High net worth companies, unincorporated associations etc.”) of The Financial Services and Markets Act 2000 (Financial Promotion) Order 2005 or is a person to whom this Memorandum may otherwise lawfully be distributed.

The information in the Materials, which does not purport to be comprehensive, has been provided by the Parties and has not been independently verified. While this information has been prepared in good faith, no representation or warranty, express or implied, is or will be made and no responsibility or liability is or will be accepted by the Parties or by any of their respective affiliates, officers, employees or agents as to or in relation to the accuracy or completeness of the information in the Materials or any other written or oral information made available to any interested party or its advisers and any such liability is expressly disclaimed.

Any prospective offering, sale and use of products and services by CurveGlobal Limited, including the use of the Materials or any offer materials made available by CurveGlobal Limited in connection with the offering and sale of products and services to or their use by audiences in the US, is subject to all applicable US Federal securities and commodities laws and regulations.

The Materials and other written or oral information supplied from time to time by or on behalf of the Parties may include statements that may constitute “forward-looking statements” within the meaning of the safe harbour provisions of the U.S. Private Securities Litigation Reform Act of 1995. Forward-looking statements are not historical facts, but instead represent only our beliefs regarding future events, many of which, by their nature, are inherently uncertain and outside our control. These statements include statements other than historical information or statements of current condition and may relate to our future plans and objectives and results. By identifying these statements in this manner, you are alerted to the possibility that our actual results and condition may differ, possibly materially, from the anticipated results and condition indicated in these forward-looking statements.

The Parties do not exclude any liability for, or remedy in respect of, fraudulent misrepresentation.

© 2019 CurveGlobal Limited.

curveglobal.com | linksriskadvisory.com 17 For more information, please contact us at +44 20 7797 1055 or [email protected]. www.curveglobal.com