SIMM Margin can be 300% higher than PB Margin for pairs trades. Below, we explain why this is, and how this will impact Hedge Funds going forward.
One of the key themes of statistical arbitrage strategies is pairs trading, which is very actively used across the Hedge Fund Industry in the Quantitative space. The pairs trade looks to find mispriced long/short opportunities in a pair of securities with the expectation of mean reversion. We have focused on Common Stock and Depository Receipt pairs of the same underlying issuer for this analysis.
The strategy funds itself, as the short sale proceeds match the cost of the long trade. The arbitrage opportunity involves taking large amounts of leverage to generate the returns required for such strategies to be profitable. This high leverage makes pairs trading a very margin-sensitive trading strategy.
ADR-Common Pairs in the traditional PB space
The pairs strategy is usually run by holding physical securities (long or short) and synthetic exposure to the securities via Equity Total Return Swaps (TRS). Importantly, these Equity TRS are in scope for the Uncleared Margin Rules (UMR).
The reason for this mismatch in instrument types is that the common stock on the local listing is in general easy to access via the physical cash market. However, the other side of the “pair”, the American Depository Receipt (ADR), is not always as liquid requiring Hedge Funds to access the depository receipts in synthetic form i.e., Equity TRS.
There is no difference from a House Prime Brokerage (PB) perspective, as the margin rates consider the pairs as the same underlier regardless of underlying product types used to access exposure. Standard market practice across the large PBs is to treat these pair trades with margin rates between 6-8% of Gross Exposure and only add a Net Exposure margin charge if the portfolio composition tilts in a certain direction.
The directionality constraints are quite tight, allowing the PB to effectively manage their risk. From the clients’ point of view, it is a pure arbitrage strategy and wouldn’t look to take directional bets keeping them in the optimal bracket to access the required leverage from their PB needed to make this strategy a success.
Impact of UMR Phase 6 and SIMM
With the lower Phase 6 (Average Aggregate Notional Amount (AANA) threshold of $8 Billion, the International Swaps Derivatives Association (ISDA) estimates an additional 775 counterparties to come into scope. In scope Hedge Funds trading Equity TRS will have to start assessing the impact of strategies where the Standard Initial Margin Model (SIMM) margin is considerably higher than House (PB).
The first point to note here is SIMM margin calculation provides a netting benefit based on risk. For example, if one leg of the pair is traded physically and the other as an equity swap, then the total margin will be duplicative (i.e. only one leg qualifies for SIMM). This is the case if only one of the legs can be accessed synthetically. Hedge Funds will have to trade both legs via equity swaps so as not to pay punitive margin rates.
The second key point is how the margin for such strategies is treated under ISDA SIMM. This has been reviewed as part of ISDA forums with the below question and response:
An ADR (American Depository Receipt) or GDR (Global Depository Receipt) is created by a global custodian who buys a number of the underlying local shares and issues depository receipts in another, second, market. Each ADR/GDR represents a fixed number of the underlying local shares. It resembles a ‘single constituent ETF’ as a closed-end fund that has only one holding. The ADR also gets its own ISIN. Should an ADR be treated as separate equity distinct from its underlying, or should it be decomposed into an FX position and its underlying equity position?
Forum decision, 23 January 2019: For simplicity, an ADR or GDR should be treated as separate equity distinct from its underlying and not decomposed into an FX position and a position in the underlying equity. This is also consistent with SIMM’s currency-neutral approach to equity aggregation.
Industry-standard is to use the International Securities Identification Number (ISIN) as the risk qualifier in SIMM, and given an ADR or GDR has a different ISIN from the common stock line, they are treated as different underliers. The fact that they are in offsetting directions and fall under the same sector buckets (e.g., large cap-developed market energy pairs will be under risk bucket 7) provides some, but not complete netting benefit in the SIMM calculation. In particular, the positions receive correlation offsets but remain as two separate positions. In essence, the pairs trade becomes no different than two unrelated positions with opposite directions in the same sector.
Now, taking the example of a large book of large market cap-developed market pairs across different sectors and individual pair trades of similar positions, we carried out calculations to produce SIMM margin numbers with the following results.
- When treated under existing ISDA guidance, SIMM margin as a percentage of Gross Exposure sits between 15-18%, for a large-cap pair from a developed market across different sectors. 3x higher than a standard PB House margin calculation.
- Taking a hypothetical assumption that the two pairs are not under two different ISINs but qualify under the same ‘Issuer’ and re-running the above analysis produced SIMM margin as a percentage of Gross Exposure between 5-0.7%. 10x slower than what a PB charges.
There is a huge disparity at either end of this spectrum. The current framework does not capture the true risk of such a strategy and is either grossly over or under-margining the strategy when compared to PB margin amounts.
The higher margin rate makes the feasibility of the pairs trading strategy a big question for hedge funds.
There may be a need to re-address this at an industry level as more hedge funds with complex strategies start assessing the impact of UMR, but until this is reviewed, it is essential that Hedge Funds that are either in scope or close to being in scope for UMR have evaluated their trading strategies under the new regulatory regime.
As demonstrated above, failure to understand this new impact on Initial Margin can result in significantly increased margin requirements, to the extent that existing trading strategies are no longer profitable.
How Cassini can help
House vs SIMM Margin- Cassini replicates internal PB house margin policies and regulatory models such as SIMM and Reg-T. Allowing you to understand opaque setups prime brokers might have. Cassini can overlay this by supporting the greater-of, distinct, and allocated approaches between Reg Initial Margin (IM) and House Independent Amount (IA).
PB Margin Replication – Cassini has a configurable framework to replicate rules-based, stress-based, and VaR-based PB Models. Clients can replicate their PB margin policies providing clients the ability to change parameters on demand in line with their brokers, compare portfolio moves between brokers and assess the impact of new mandates.
Pre-Trade What-If Analysis – At the time of execution, you can access Cassin analytics to understand what the impact of the trade will be against not only the chosen counterparty but all other routes to market giving you the ability to allocate most cost-effectively.