Trading Strategy as a “Shadow” Part of Market Risk

When considering the contribution of a trading strategy to the overall risk of the position, we need to examine what determines a trader’s attitude to risk. Admittedly, traders have a greater participation in the upside than in the downside of their trades (Allen 2003). This means that a typical trader’s compensation resembles a payoff on a longcall option with bonuses linked to profit which are potentially unlimited, while the financial share a trader bears in losses is capped at his salary and any deferred payments. Some of the highest disclosed traders’ compensations are a good illustration of the upside potential, for example; Driss Ben-Brahmin (Goldman Sachs) reportedly earned about £30 in 2006 (BBC 2004), Brian Hunter (Amaranth) received over $100 m in 2005 (Petzel 2006), and Adam Levinson (Fortress) was remunerated with £156 m in 2008 (Antonowicz 2008).

Since vega of a long option position is positive, traders have strong incentives for risk-loving behavior because it increases their expected payoff. As long as a profitable trading strategy keeps producing alpha (and, consequently, bonuses for its owner), traders are reluctant to share the details about their strategies and risks they are about to take with their peers and risk managers.

This information asymmetry is inherent to any financial firm and means that the risk of individual trading strategies and their undesirable interactions may not be properly detected and controlled by the firm’s risk managers. This, in turn, enables rogue traders to take and pile up hidden risks that, if realized, may increase the trader’s propensity for operational risk, as illustrated by the collapse of Baring’s.

Ironically, losses incurred as a result of rogue risk-taking do not necessarily mean an end of the trader’s career. Some evidence indicates that the trader’s market value may be negatively correlated with failures. According to Allen (2001), “Even firing does not have that large an effect – the tendency is for firms to hire traders who have had spectacular blowups elsewhere, figuring they’ve learned a lesson (at someone else’s expense). NickLeeson going to jail was an aberration (possibly due to different attitudes in Singapore than in the West).” The years which have passed since the demise of Baring’s have shown that Mr. Leeson’s case was a precedent rather than an exception. Since Baring’s collapse, the rogue trader “hall of fame” has expanded to include Yasuo Hamanaka from Sumitomo (8 years in jail following a $2.6 bn loss on copper trades in 1996), John Rusnak from the Allied Irish Bank (7.5 years in jail after $691 m loss on FX options), Brian Hunter from Amaranth (hedge fund liquidated after a loss of $6.69 bn on natural gas futures), Jerome Kerviel from Societe General (5 years in jail following a loss of €4.9 bn), and Kweku Adoboli from UBS (7 years in jail after a loss of $2.3 bn on stock index futures).[31] In hindsight, risk management in these institutions should have been held responsible for failing to prevent these losses.

The trader’s risk appetite can be curbed by means of more symmetrical com­pensation schemes (e. g. ‘golden cuffs’ and cash bonus clawbacks), internal controls (e. g. regular audits, phone conversation recording), or even pre-committing traders to specific loss limits by incentivizing them to share their forecasts with risk managers[32] (Miller 2001). It is more common, however, to limit the risk traders may take from the top-down rather than bottom-up, by having a trader stick to externally set limits. A typical market risk limit structure in a financial firm includes various position limits, P&L (stop-loss) limits, limits on specific risk parameters (e. g. rate buckets, “the Greeks,” markets, and liquidity), total risk limits (VaR/CVaR limits), and limits based on stress testing. Now we will consider more closely the interplay of risk limits, with VaR as a risk measure, and exposure (position) limits.

Leave a reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>