Margin notes

Despite being laid off from his job at a power trading company, Joe Risk Manager is feeling pretty good. He has landed on his feet as the new chief risk officer for a small oil producer called Risky Oil. The firm can produce 10 million barrels of oil a year, but operates on very thin profit margins, producing at an average cost of $23 a barrel (bbl). In addition, the company carries significant debt. Given this situation Joe helped the company implement a policy of hedging its production to stabilise cashflows.

Specifically, Risky Oil hedged its expected production for the next five years with a swap negotiated with an energy trading company. The swap guarantees that Risky Oil will receive a fixed price of $25/bbl for the next five years. Mindful of the credit risk of such a long-term contract, Joe also helped negotiate a collateral agreement between the two counterparties, which requires both firms to post either cash or a letter of credit equal to any mark-to-market (MtM) loss on the position.

As a result, Joe has helped manage both the market and credit risk of the company. However, he shouldn’t get too comfortable – he may still be sitting on a time bomb.

The problem is that while Joe has managed the market and credit risk, he has not considered margin risk. Margin risk refers to the amount of cash that a company may be required to raise to meet margin calls based on MtM losses in the position it has taken. If oil forward prices rise, Risky Oil would be contractually required to post substantial collateral to cover the loss to the energy trading company.

For example, if the price of oil underlying the hedge were to rise dramatically to $30, Risky Oil could be asked to post $250 million in collateral¹.

However, Risky Oil does not have much cash and, given the current debt level, banks may not be willing to issue additional letters of credit. The unfortunate result of this chain of events is that Risky Oil, a company that on paper has not lost any money, may be forced to declare bankruptcy. While this might appear a theoretical example, margin risk is very real. In fact, it has contributed to some of the major trading disasters in recent years including those at Metallgesellschaft, Ashanti Goldfields and Long Term Capital Management.

Measuring margin risk
So how can we manage margin risk? The first step is to measure it, and unfortunately this is where most analysis stops. We need a substantial amount of data to calculate margin risk. Specifically, we need to know: positions by counterparty, the specific margining rules by counterparty (including credit limit, type of collateral required and netting agreements²), current MtM by counterparty, current collateral by counterparty, market prices, volatilities and correlations.

The difficulties associated with gathering this data often keeps risk managers from calculating margin risk. Yet if we can collect this data set, we can use a Monte Carlo simulation to measure margin value-at-risk (Var). The steps to take are as follows:

1. Generate a price scenario.
   
2. Create a new mark-to-market value for each counterparty based on the price scenario, positions with the counterparty and current market-to-market.
   
3. Calculate the collateral required for each counterparty.
   
4. Determine total collateral requirements taking into account potential rehypothecation³.
   
5. Compare simulated total collateral requirements to current collateral requirements. This is the change in margin.
   
6. Repeat steps 1–5 thousands of times.
   
7. Determine the change in margin associated with the 95% level from all scenarios. This is the 95% margin Var over the selected time horizon.

Simply making the effort to calculate this measure has significant value. But that is only the first step. Margin risk has attributes of both market and credit risk. In some situations, we can hedge margin risk in the same way we could hedge market risk, although most of the time margin risk is more similar to credit risk and very difficult to eliminate (see box and table). For this reason, the margin Var number should be calculated over a number of different time horizons. Only if we do this will we understand how much cash a company may be required to post as margin.

We must bear in mind that margin Var is a fundamentally different measure from a measure of market or credit risk. This is because margin risk does not represent a true financial loss, such as market or credit risk, but rather a use of capital. If capital is unconstrained, there is no reason to worry about margin risk. Given that margin risk looks at uses of capital while market and credit risk examine financial loss, margin Var should be calculated separately from market or credit risk measures.

These measures of margin risk tell us how bad our positions could get under the current margining rules – a valuable piece of information. However, given events in the energy industry over the past year, another important measure is the impact on margin requirements of a credit downgrade. To measure this impact, we need to also know how the margin requirements for every counterparty depend on credit rating.

For example, consider a company with a $25 million MtM loss on a position with a counterparty. If, based on its current credit rating, the counterparty grants it a credit line of $20 million, the company would only have to post $5 million in collateral. Yet if the firm were downgraded, it might only receive a $10 million credit line and be required to immediately post $15 million in collateral.

Hence, if we know such parameters by counterparty, it is a simple matter of recalculating a margin Var under the downgrade rules to determine the impact of any downgrade.

Managing margin risk
Once we are able to measure margin risk, we can try to manage it. The first step is taken during negotiation of the credit agreements with counterparties. Ideally, we would always prefer to receive collateral and not be required to post any. But few – if any – market participants are in a position to demand such terms.

Alternatively, we can focus on obtaining collateral that can be rehypothecated or at least on insisting on cash collateral beyond a certain threshold. Another possibility is negotiating higher credit lines. However, obtaining them may require giving the counterparty higher credit lines as well, thereby increasing potential credit risk.

Even after a company has negotiated all credit agreements, it can still manage margin risk. One method is to use a clearing service. By netting positions through a single counterparty, we can dramatically reduce the amount of collateral required and thus margin risk. However, while numerous suggestions have been made for the creation of clearing-type organisations, there has been no clear emergence of a standard in the energy industry.

In the absence of clearing, we need to evaluate the impact of trades before they are executed. If there is a choice of counterparties for a trade, then, everything else being equal, we would prefer to make the transaction with the counterparty that causes the smallest increase in our margin Var.

To provide traders with the proper incentives, we could actually charge a trader proportionally to the impact of a trade on margin risk. If the trade reduced margin risk, traders would receive a credit for that trade. Of course, we must bear in mind that margin risk is only a small portion of all the risks associated with a trade. The ideal system would give traders a charge that incorporates the marginal impact of the new trade on market, credit and margin risk.

The risk manager’s role has changed dramatically over the past 10 years. Historically, we have focused on the big risks associated with market and credit losses. These risks are now well monitored and well managed. As we go forward, the risks that threaten companies are likely to come from elsewhere.

Margin risk is one such example. By monitoring and managing this risk as well, the risk manager will – hopefully – prevent his company from joining the ranks of firms that have failed due to unexpected margin risk.

Brett Humphreys is a consultant at Risk Capital Management in New York
e-mail: humphreys@e-rcm.com