Commodities threaten pain for initial margins

Market volatility and spot-based Simm approach may drive large margin spikes in little-tested asset class

  • On September 1, an estimated 800 firms will be caught in scope for the final phase of the rules on non-cleared margin. The majority buy-side cohort brings with it an array of trades that have been largely untested in the regime thus far.
  • Commodities pose a particular challenge as price volatility could cause margins to spike and potentially breach regulatory relief thresholds.
  • For commodities, the standard initial margin model (Simm) differs from central counterparty models, as it ignores contract tenors, instead using spot price as the sole risk factor. Some worry this does not reflect the way commodities are traded.
  • Although the focus on spot increases netting, it also sees exposures margined at the most volatile part of the curve and could exacerbate spikes.
  • Some vendors are struggling to access the required data for Simm calculations and backtesting.

One of the most tangible impacts on the global economy of Russia’s invasion of Ukraine has been the disruption caused to supplies of oil and grain. This has led to a historic hike in commodity prices – Bloomberg’s BCOM index is 28% up for the year to date – that is also posing challenges to efforts to regulate derivatives.

The sixth and final phase of the implementation of non-cleared margin rules, which comes into effect on September 1, will affect an estimated 800 firms. These businesses will bring with them an array of complex portfolios and exposures that have been largely untested during the previous five phases of the rules’ implementation.

Chief among these exposures are commodities, where data required to calculate initial margin can be scarce. Margin amounts for commodities can also spike due to whipsawing prices and a calculation model that assumes all contracts are traded on spot.

“With margin, we tend to see the most volatility in commodities,” says Jo Burnham, risk and margining expert at OpenGamma. “It’s prevalent in equities too, but they tend to be more spread around, so you only get big jumps in certain products.

“Commodity portfolios can be very directional because it’s against the actual physical, and is also quite long-term. There’s usually a maximum amount that interest rate products can move by – unlike commodities, where you can have sudden huge changes in mark-to-market.”

That could spell danger for more than three-quarters of phase six firms that are expected to take advantage of regulatory relief. This exempts counterparties from compliance if margin exchange amounts remain below $50 million.

Vendors have noted a surge in commodity trades as hedge funds and commodity trading houses prepare to fall within the scope of the new requirements. Many of the firms subject to phase five’s implementation last September have also begun to add more complex commodity exposures in response to rising asset prices.

“Recently we’ve seen an uptick in products like commodity quanto swaps, mostly from Asian clients on the sell side,” says Eduardo Pereira, product manager for standard initial margin model (Simm) at Bloomberg. “Phases five and six also bring more hedge funds into scope, and they have a lot of commodity exposure because the high volatility in commodity markets over the past year, and their function as a natural inflation hedge has brought a lot of trading interest. 

Scott Fitzpatrick, chief operating officer at risk technology vendor Acadia, agrees that commodity portfolios are more commonplace in the 600 or so clients it expects to onboard for the final phase.

“This time we are seeing more commodities and we’re seeing them in all regions,” he says. “There’s definitely a larger commodities footprint in phase six.”

On the spot

Commodities are an outlier when it comes to calculating margin under Simm. The model, developed by the International Swaps and Derivatives Association, calculates regulatory initial margin based on the risk sensitivity of portfolios and typically leads to lower margin amounts for balanced portfolios compared with regulators’ standard grid methodology. Simm has been adopted by most of the firms that fell under the purview of the earlier phases of the rules.

Under Simm, most asset classes have different risk factors depending on the tenor of contracts: interest rates, for example, have 12 different risk factor tenors, while credit has five. Commodities attract different risk weightings depending on the asset, ranging from 13% for some agricultural products to 53% for freight derivatives; the weighting for crude oil is 29%. However, the only risk factor is the spot price.

“UMR margin assumes the whole of the risk is on the spot price of the commodity, but that’s not necessarily the way it’s traded,” says OpenGamma’s Burnham. “A lot of oil trading is almost like interest rate trading as it’s on a curve. If you look at metals, they’re often out to long dates, but that’s not how Simm or grid look at it.”

The single risk factor leads to greater netting. This means Simm typically delivers lower margin amounts than the grid, which focuses on gross notional.

However, the model differs from those used in exchange-traded markets, which incorporate multiple tenor risk factors. For example, the Span models currently used to calculate initial margin on commodity contracts at CME and LME do account for contract tenors.

Commodities can deliver some unusual results under Simm. Analysis by OpenGamma shows that a crude oil basis swap with equal notional on the two legs would attract 15% margin under grid but only 4% under Simm. However, the industry model can be more punitive for some portfolios: a swap, with the same notional, on long middle distillates and long European natural gas would attract 28% margin under Simm and as little as 12% under grid.

Some are critical of an approach that focuses only on the most volatile part of the curve. “For commodities, Simm focuses too much on spot risk,” says Andy Shaw, chief executive of advisory firm Links Risk. “When you look at the extent of the backwardation in some of these markets, that’s a weakness if you have a portfolio exposed to a single commodity.”

The breakdown in global supply chains has flipped curves into deep backwardation. This means longer-dated contracts trade cheaply compared with near-term contracts and spot. In normal times, physical commodities tend to trade in contango, with longer-dated contracts pricing above spot to reflect the cost of storage.

Tara Kruse, global head of data, infrastructure and non-cleared margin at Isda, says that Simm responds to the regulatory requirements for all material risks to be modelled.

“The experience of the Isda Simm is that commodity term structure is not usually a material risk in derivatives portfolios, even for those portfolios that are concentrated in the commodity asset class,” she says. “Consequently, commodity term structure risk is not modelled in the Isda Simm.”

She adds that this is also the case for regulators’ own sensitivity-based approach for the Fundamental Review of the Trading Book, on which Simm is based.

Isda monitors the model’s performance on an annual and quarterly basis. “We would be able to detect if there were cases of significant under-margining due to this risk factor,” Kruse says. “If that were to happen, then we would be ready to take action to remedy the situation.” Kruse says she has seen no instance of such a problem since the model became operational in 2016.

Links Risk’s Shaw believes the introduction of smaller, more concentrated portfolios could deliver more backtesting failures than were witnessed during previous phases, where diversification may have kept some weaknesses hidden from view.

Regulators require firms to backtest their regulatory initial margin models – including Simm – to ensure the businesses can cover 99% of 10-day close-out losses. Isda recommends a three-plus-one approach – comprising the last three years plus a stressed year, which is deemed to be the 12 months covering the 2008-09 financial crisis.

“You only really find the problem with Simm when you go in and forensically scrutinise it with backtesting,” says Shaw. “With phase one firms you get so much diversity in their portfolios that Simm backtests very conservatively and looks to be a significant overestimate with respect to regulatory defined thresholds. This has masked some of the quirks of the model that could cause problems for smaller portfolios.

“However, you’ll see more examples of it not working so well as it is applied to these smaller, more concentrated portfolios. You’ll see that across other asset classes as well, but commodities are in the spotlight now.”

He says the kinds of trades that may fail Simm backtests include large curve positions with distant tenors – such as trades that are long one-month spot risk and that offset short spot risk in 10-year contracts: “You tend not to get so many of those idiosyncratic portfolios, but you’ll get more of them in phase six as smaller relationships come on board. It’s going to be a problem from time to time.”

Where portfolios fail a backtest, firms must switch to the grid approach, use an alternative backtest methodology or provide an additional margin buffer to cover any shortfall.

Data dilemma

One of the major hurdles for backtesting, particularly for complex products, is sourcing data going back more than a decade.

“You need to know what your risk factors were for these commodity products back in that financial crisis period of 2008-09,” says Pereira. “That is something where we see market participants struggling across the board – actually having the data and the risk factors to compute P&L on these historical dates, so you can provide evidence to your regulators that you have sufficient margin to cover for the P&L in your portfolio. That is absolutely a point of debate in the market.”

With more than 40 licensed Simm vendors, some are struggling to access the data required to calculate the commodity risk sensitivities that are needed to support clients’ margin calculations.

“You’d think you get to phase six and all the products would have been covered,” says Craig Pearson, director of Margin Tonic, a consultancy. “But as we’re getting into the details of some of these nuanced commodity products, there’s quite a lot of activity on behalf of vendors to understand those products and get access to that market data to price those products and define their models. Vendors typically build as required, so if their clients haven’t had those products before, they won’t have developed them.”

Some vendors had a head start. Acadia, which is already the largest provider of Simm services, says it is up and running with commodities after onboarding some large players in the asset class during earlier phases.

“Some firms which were active in that space used us before, which was a big test of our coverage and confirmation that our models lined up the way clients were expecting,” says Fitzpatrick.

Others, such as Bloomberg, benefit from owning much of the in-demand data. “We have quite complete datasets from a volatility surface point of view, which is important to price options on commodities,” says Pereira. “That’s where the market is struggling and where we can add some value in the completeness of our Simm solution.”

He adds that generating Simm for complex products – such as commodity quanto swaps, where asset prices are expressed in a currency other than the one used for pricing – call for an array of risk factors to be computed. This can only be done with a large library of pricing and valuation data.

“To generate Simm on commodity quanto swaps, there are certain risk factors you have to compute, and we’re beginning to see more appetite for these structures, particularly in Asia,” says Pereira. “It can cause a bit of a headache to really understand the risks they should generate from a Simm point of view and it’s clearly something we’ve seen with some sophisticated banks coming into scope right now.”

Margin Tonic's Pearson says that although the required data does exist, it is not always available to vendors. Clients may source the necessary data for pricing and valuation purposes, but licensing terms do not always permit this to be passed on to third parties.

“The market data exists because anyone trading these products must inherently be able to price them," says Pearson. "However, where firms are using a Simm vendor, a classic issue is where the client has access to the necessary data, but the terms of the licence prohibit them from passing data on to a secondary user. Simm vendors must therefore identify their own data source or work with their clients to resolve licences to enable them to access the newly required data."

Editing by Daniel Blackburn

  • LinkedIn  
  • Save this article
  • Print this page  

Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.

To access these options, along with all other subscription benefits, please contact [email protected] or view our subscription options here: http://subscriptions.risk.net/subscribe

You are currently unable to copy this content. Please contact [email protected] to find out more.

You need to sign in to use this feature. If you don’t have a Risk.net account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an individual account here: