Hedge funds make use of leverage in order to optimise their returns. They use a combination of derivatives, where they need only fund the margin and not the total value, and cash trades funded through loans from a prime broker, to maximise the potential profits on the cash available to them.

One of the most popular trading strategies is the Treasury basis trade. This involves taking a short position in Treasury futures and borrowing money to take a long position in cash Treasuries. But these trades are no longer as efficient as they once were, with a combination of high margin levels and high funding costs impacting the available leverage.

In 2008, firms were able to achieve high levels of leverage on their trading. But then the global financial crisis, following the bankruptcy of Lehman Brothers, changed everything. The new regulation, such as Dodd-Frank and EMIR, introduced in the aftermath of these events, forced an increasing number of firms to post collateral for their derivatives positions.

The increase in the range of derivatives that are now cleared and the amount of collateral that firms have to post has highlighted the issue of fragmentation. Exchanges are inherently regional and product specific, meaning that offsetting positions will be margined separately.

The portfolio-based models used to calculate Initial Margin, such as the older scenario-based algorithms like SPAN and IRM 1, or the newer VaR-based models like SPAN 2, IRM 2 or Eurex Prisma, will net potential profits and losses under their various scenarios for positions that are cleared under the same CCP, but for other parts of the portfolio a completely separate Initial Margin will be calculated, leading in effect to gross margining.

This fragmentation led to a decrease in the leverage available to firms, with many only able to achieve half the previous levels. But that was before the additional volatility of the past few years.

It’s been a rocky ride, including Covid, the Russian Invasion of Ukraine, the disastrous UK mini budget and some major weather events. The resultant price volatility has led to significant increases in margin requirements, with hikes of over 100% in a short period of time not unusual – which had a significant impact on the amount of leverage firms were getting from their derivatives positions

Although some of the volatility has gone, margins have remained higher than the historic norm. The current geopolitical situation and the impact of climate change means that it is likely that events could easily lead to a return to low leverage from derivatives. And that is without the introduction of mandatory clearing for repo and treasury that will directly impact the Treasury basis trade.

How to solve the leverage problem

The recent past has left derivatives users with a difficult decision:

  • Try and maintain leverage without sourcing any additional assets, leaving yourself exposed to critical liquidity risk.
  • Reduce leverage  by increasing assets, but this will be a drag on returns.

However, there is an alternative. Firms still need to decide the amount of liquidity risk they are willing to accept, and use stress testing to size their funding requirements appropriately. Without any other action, the funding level will be based on their current portfolio and the way that margin is calculated on this. But there are ways to optimise margin requirements to maximise the available leverage:

Same product using alternative venues

There are lots of examples of the same product being available on multiple markets:

  • Major currency swap clearing is available across multiple exchanges. In particular USD and EUR swaps out to 50 years can be cleared at LCH, CME and Eurex
  • TTF gas futures are available on ICE, EEX, Nasdaq and CME, although the contract specifications vary (for example, some are deliverable whereas others are cash settled).
  • Brent and WTI futures are available on both ICE and CME, although the level of open interest varies.

Given that the products are the same and that all the clearing houses are calculating margin based on the same or similar regulation, it is surprising how different the requirements can be.

Example

The following table shows, as percentages of notional, example margins calculated on a series of par swaps at the major CCPs clearing interest rate swaps.

EURCCP1CCP2CCP3
Tenor (Years)PayReceivePayReceivePayReceive
20.46%0.45%0.44%0.26%0.39%0.37%
51.29%1.58%1.24%0.96%1.07%1.16%
102.59%3.05%2.74%2.03%2.13%2.78%
205.73%4.85%5.47%3.85%6.30%5.28%
309.31%6.48%9.64%6.71%10.98%7.27%
5014.01%10.40%18.84%10.44%18.28%12.43%

It can be seen that where CCPs stand in relation to one another is dependent on the tenor of the swap. For example, CCP3 margin is less than CCP2 for a 2 year swap, but more than CCP2 for a 20 year swap.

And looking specifically at the margins calculated, for a EUR 100 million notional 20 year swap this could mean a reduction in margin of 14 million Euros clearing through CCP2 rather than CCP3.

Same risk using alternative products

There can be different ways of trading the same level of risk that can have a significant impact on the level of margin required or the timing of any cash flows, both of which feed into liquidity requirements:

  • Different product types can be used, for example, options can be used to simulate a future position.
  • For many underlyings there is a choice of cash settled or deliverable derivatives.
  • Many products can be traded both cleared and bilateral.

For these alternatives sometimes the difference in margin is small, but at other times it can be massive.

  • For some bilateral trades no margin will need to be paid making it a significantly cheaper option.
  • However, if the trade is subject to UMR, unless you are using the $50M threshold, then the margin required can be a lot higher than that calculated for cleared trades. Differences of over 4 times are often seen.

Example

FX options are often traded bilaterally, but can also be traded on exchange, for example at CME. For the bilateral positions the margin may be calculated using either SIMM or Grid, whereas on CME, SPAN is used. The following table shows the difference in requirements for various FX option strategies:

StrategyMargin
SPANSIMM
Butterfly  36,410.00195,992.18
ATM Call164,230.00381,269.99
Straddle  86,680.00429,264.48
Straddle Spread  31,364.08  98,549.81

As can be seen, the SIMM margin is significantly higher, being over 5 times as much as SPAN for the Butterfly strategy.

Same portfolio using alternative clearers

Choosing where to place your cleared business can have a significant impact on the margin charged. Optimising the level of broker costs is obvious. Each will have its own schedule of charges and may also use their own margin algorithms.  But if they are replicating the CCP margin calculations they will also have varying multipliers that will be applied.

Outside of the use of different algorithms or multipliers, there are two ways that choice of broker can impact the actual margin level:

  • Taking into account existing positions held with the broker is important so that any offsets can be maximised.
  • But the biggest savings can be made by splitting positions between brokers to minimise the liquidity charge.

Example

With the introduction of the non-linear liquidity add-on, the traditional way of allocating business between clearing brokers, putting all positions from one market at the same broker, is now suboptimal. Any large outright position will incur significant margin in order for the CCP to cover the risk of close out on default. If you allocate your trades by market it may give the maximum position offset, but also maximum liquidity charge.

Assume you have the following Eurex positions – a combination of fixed income and equity, with the fixed income positions allocated to broker 1 and the equity positions allocated to broker 2:

InstrumentPositionBroker
March Bund 164 Call OptionLong 1,000Broker 1
March Bund 163 Put OptionLong 1,000Broker 1
June DAX 16500 Call OptionLong 1,200Broker 2
June DAX 16000 Put OptionLong 1,200Broker 2

The initial margin calculated will be as follows:

VaR Margin ComponentLiquidity Add-onTotal IM
Broker 1998,305531,5181,529,823
Broker 25,316,9152,367,6347,684,550
Total6,315,2202,889,1529,204,372

A more efficient way of allocating positions needs to be found which will minimise the margin calculated by clearing brokers.  This means maximising the offsets between contracts and minimising positions for a given product allocated to a broker.

A simple split of the positions between the 2 clearing brokers could be as follows:

InstrumentPositionBroker
March Bund 164 Call OptionLong 500Broker 1
March Bund 163 Put OptionLong 500Broker 1
June DAX 16500 Call OptionLong 600Broker 1
June DAX 16000 Put OptionLong 600Broker 1
March Bund 164 Call OptionLong 500Broker 2
March Bund 163 Put OptionLong 500Broker 2
June DAX 16500 Call OptionLong 600Broker 2
June DAX 16000 Put OptionLong 600Broker 2

This will result in the following initial margin:

VaR Margin ComponentLiquidity Add-onTotal IM
Broker 13,157,610764,6443,922,254
Broker 23,157,610764,6443,922,254
Total6,315,2201,529,2887,844,508

This is a margin saving of EUR 1,359,864 or 15%.  And if there were four clearing brokers available the total margin would be reduced to EUR 7,159,572 which is a 22% margin saving.

Conclusion

Since 2008, the level of leverage available when trading derivatives has declined considerably. From an average of around 10x leverage, a combination of regulation, fragmentation and volatility has severely reduced this. At its worst only 0-2x leverage has been achievable.

If firms want to continue to benefit from leverage and not let margin be a drag on their returns then they need to look to optimise their requirements. Making the right choice of where and what to trade can have a significant impact on margin, allowing firms to limit the amount of funding required without exposing themselves to excessive liquidity risk.