What commodity risk managers can learn from JP Morgan's $6bn loss
What was the London Whale?
In the first half of 2012, a wide circle of hedge fund managers, credit industry journalists and the general finance community became aware of a substantial loss-making position held by JP Morgan & Co. The company held positions in credit default swap indices linked to the default likelihood of US corporates. By mid-2012 the estimated loss on the position was $2bn but this was revised upwards to $6.2bn. The loss could have been greater if was not for the calmness of the wider economy at that time. Prior to this JP Morgan was on-track for a $10bn net-profit in 2012.
Due to the sheer size of the loss, the position became known as the London Whale. It is remembered as one of the biggest single losses in financial history.
As of 2011 JP Morgan’s Synthetic Credit Portfolio (SCP) generated circa $2bn in gross revenue for the firm. The portfolio traded credit default swaps (CDS). CDSs are a type of derivative that act as an insurance contract where the buyer is compensated by the seller in the case of a loan default. The overall CDS market capitalization was on the decline since the 2007-2010 recession reaching $25.2trn by 2012 from $62.2trn in 2007.
CDSs are not exchange traded. They are exchanged over the counter (OTC) between counterparties. This means that their pricing is cloudier than other derivatives. Analysts became concerned over the size and opaqueness of the CDS market during 2007-2010 financial crisis with some suggesting CDSs caused market defaults due to moral hazard (i.e. counter parties receiving more from default protection than they would lose from the default itself).
JP Morgan used the activities of the SCP to offset, i.e. hedge, the credit risk the wider firm faced as a large-scale lender. It also partly used the SCP to generate trading revenue. In 2011 JP Morgan had an improving economic view after activity began to pick-up at the end of the 2008-2010 recession. This meant the bank believed it required less credit default protection. As such management and traders of the SCP were tasked with scaling back the riskiness of their CDS investments.
At this point the situation began to unravel.
Initially the SCP traders looked to close-out long and short positions to reduce risk. This, however, proved to be costly, with a predicted $500m total loss from unwinding 25-35% of the SCP’s positions. According to JP Morgan’s internal report, SCP management and traders asked for and received permission to instead increase their positions in certain types of CDS indices to offset existing positions. They were looking to reduce risk by hedging and in doing so they increased the size of the SCP portfolio from $50bn in December 2011 to $157bn in March 2012. However, the SCP was using financial models built into Excel spreadsheets that assumed correlation between longer-term CDSs and shorter-term CDSs. This correlation was necessary for using the longer-term assets to hedge the shorter-term ones. In fact, this correlation had been declining to a point where their hedging strategy was no long effective. This meant the bank’s existing valuation spreadsheets were no longer fit for purpose.
Compounding this problem were several managerial and monitoring errors. For several months, the traders were allowed to choose which part of the daily intraday price range they marked their positions to. This essentially meant that the traders were “marking their own homework”. At least once, traders waited until 20:00 London time to mark their trades in the hope of getting a better price during New York trading.
Meanwhile, as the SCP traders continued to build their positions, the head of JP Morgan’s Chief Investment Office – tasked with overseeing the SCP – did not receive full position reports from the portfolio. As a result, they did not fully understand the size of the CDS portfolio. Instead they received profit/ loss statements and aggregated risk metrics. The profit/ loss statements were flawed due to financial modelling weakness and traders choosing preferable pricing. The aggregate risk models hid the SCP portfolios risk within the bank’s wider investment portfolio. Finally, not only were scheduled risk management meetings poorly attended and often delayed but the SCP’s Value-at-Risk (VaR) limits were breached more than 300 times in early 2012 - until a new, more lenient VaR model was put in place. The new VaR model cut the initial loss estimates from the SCP’s CDS positions by half, emboldening traders to further expand their positions.
The London Whale trade led to billions in losses for the bank, the resignation or termination of multiple employees and heavy fines from regulators.
What can treasurers and risk managers learn from the London Whale?
In early 2013, JP Morgan published an internal task force report detailing the errors that lead to the losses. The management decisions and indecisions, weak internal reporting and weak risk management behind the loss provide many lessons for treasurers and risk managers to take on-board.
The London Whale case may seem part of the strange and esoteric world of high stakes finance. If you may think the events that led to such a substantial loss can never happen in your working world, then think again. Any company that faces a volatile market price as a major input cost is at risk of stumbling down JP Morgan’s path. And any company that lets its guard down can face another London Whale.
So what lessons can commodity risk managers learn here?
1. Ensure a complete and holistic view in reporting
One of the key failures of JP Morgan’s risk management was the reduced and distilled view the Chief Investment Officer (CIO) had on the SCP’s CDS portfolio. The information received was limited in scope and the CIO did not see a full position report from the SCP. On the other end, the CIO was selective in reporting the activities of the SCP and what little information they had on the SCPs portfolios to JP Morgan senior management. Information on the positions were being filtered and clogged up in reporting chains. The wider business did not understand the risk until it was too late.
A commodity risk manager should work to ensure there is ‘one source of the truth’ when it comes to commodity volume forecasts, fixed price volumes, mark-to-market calculations and risk metrics. All the key stakeholders should ‘sing from the same hymn sheet’ which should be prepared with a clear process that reduces human interference.
2. Make management information available to all that need it
Information about a business’s commodity risk and hedging risk should never stay within the specialist team responsible for managing that risk. Multiple business areas from finance, to treasury, to procurement and upwards to the company’s senior management must have an unfiltered and clear view of the company’s commodity positions and risk.
3. You need rapid access to position reports
Wherever possible, technology should be leveraged and processes put in place to ensure frictionless, accurate and timely flow of information from the various business units. Markets move quickly – position and risk reporting should match their tempo.
4. Controls and risk monitoring:
a. Be granular in your analysis
b. Limits are limits (they are not there to be broken)
c. Ensure your risk managers have effective authority
JP Morgan erred in two ways with its risk monitoring and control protocols in this situation. Firstly, VaR limits were not hard limits. Hundreds of VaR breaches were ignored and a new untested VaR model was put in-place which ultimately failed to prevent loss. Secondly, their risk metrics were not granular enough. Risk metrics were reported at the CIO level, which included aggregate risks from multiple other portfolios. This drowned out the risk present in the SCP portfolio.
It is easy as a commodity risk manager to allow a VaR or other risk metric to breach a few times. Nine times out of ten nothing will come of it. The company will face no major loss. It is much more difficult for the commodity risk manager to be firm on risk limits and push back against hedgers looking to gain a little extra breathing room as they bump up against their risk limits. Especially if they lack the bark and bite from company policy to do so. However, every now and then a risk limit will be breached and the breach will go unattended to and this will lead to substantial loss to the business. In the desert it never rains, until that one day when there are torrential floods that you did not prepare for. The appropriate course of action is to ensure risk policies provide the authority and mandate for risk managers to enforce risk limits. A commodity risk manager should have board provided authority to enforce risk limits. They should be prevented from suspending those limits on their own volition.
It’s also easier to view risk as an aggregate figure either with all the company’s commodities lumped into one group or even with commodities and other market risk lumped together. It provides a single number for management and simplifies calculation and reporting requirements. However, each commodity market the company is exposed to will have its own dynamics and risks. Hidden within that single number may be a substantial and growing risk from one commodity. An aggregate number may make you feel safe, but there may be obscured dangers. Modern corporate management favours ‘one-pagers’ and ‘KPIs’, unfortunately chaotic markets do not. To ensure no risks are concealed, it is best practise to report on and manage both aggregate risk and individual commodity risk.
5. Make sure the relevant skills and know-how are in place outside your trading team
As the SCP traders were building up the CDS positions, the SCP risk management department of JP Morgan’s was understaffed with insufficient leadership. In February 2012, a new risk manager was hired. However, they lacked any previous expertise in capital market risk management. Moreover, JP Morgan did not have sufficient human capital with experience in CDS market risk in 2011 and early 2012 to support the new risk manager. A critical watchdog function lacked bark and bite to challenge the SCP’s trading activity. This meant that the know-how on CDS markets was concentrated within the team that would be rewarded for making profit on the portfolio.
A similar story can be told in many corporates with commodity procurement and hedging teams. Specialist knowledge of commodity derivatives, commodity market terminology and commodity market risk is concentrated within these departments. The staff who will be rewarded for achieving beneficial commodity input prices know the most about commodity markets. The staff tasked with ensuring the company is protected from commodity market risk often know little of how these markets work. They too often rely on the commodity procurement teams for knowledge.
Companies should hire commodity risk managers who have strong skill sets in commodity markets or provide good, independent training for staff elevated to that role. They should ensure their compensation is not linked to getting the best market price. And they should ensure that there are subject matter experts in commodity risk outside of commodity procurement teams either internally of via external consultants.
6. Avoid conflicts of interest - no one should mark their own homework
The clear conflicts of interest in the London Whale story are surprising given how much regulatory scrutiny large banks are under. The SCP traders at JP Morgan could mark their positions against the market quotes of their choosing in a trading day. They were responsible for informing their management and the wider business on the value of their positions. The traders were also not clearly told by JP Morgan management that they would be fairly compensate even if the SCP faced losses.
Lastly, the CIO risk manager appointed in early 2012 had a ‘dotted line’ reporting structure to the head of the investment office – meaning they were managed by the person they were supposed to monitor. Each of these presents a different type of conflict of interest. These conflicts of interests will distort any risk management strategy. They appear to have exasperated the risk management failings at JP Morgan, added to the loss the firm faced and delayed an adequate response to the situation.
A commodity risk manager should not allow a commodity procurement team to be the sole source of information on the value of hedging positions. They should ensure an independent source of market information is available to mark the company’s positions against the market. This reiterates lesson 1.
From time to time commodity procurement teams will need to conduct risk trades. They will need to hedge or unhedge positions due to the breach of risk limits. It should be made clear that they will be compensated fairly for reducing the firms risk and the profit or loss on these trades will have no impact on their remuneration.
Lastly, embodied within a business’ commodity risk policy should be a clear separation of reporting lines between those who manage the company’s commodity risk and those who monitor them.
Actively managing commodity price risk is becoming more and more prevalent, demanded by stakeholders and necessary. Getting it right can be a source of significant value – but as the London Whale story demonstrates, getting it wrong can be equally costly. The good news is that there are well established governance and compliance procedures which, if implemented and monitored correctly and with the use of the right technology, can ensure the upside is maximised, and any downside minimised.