Gamma’d by the Swiss Central Bank (SNB)
The Swiss central bank (SNB) suddenly abandoned its peg of the SFr against the Euro on 15 January after having promised only 3 days before that it would not remove the peg.
The graph below shows what happened immediately after the announcement at 9.40 am on 15 January. The SFr jumped within minutes by nearly 39% and within a few more minutes settled at slightly below 1.04 SFr to the Euro. This scale of sudden jump has never previously occurred in a major currency.
Figure 3: SFr movement in the hours surrounding the SNB announcemen
Prices in financial markets are volatile, but financial market participants are accustomed to that volatility and they act in a way that ensures they survive it. For instance, the 50% drop in the oil price between between early October and January is causing grief for energy producers but it did not cause catastrophic losses for traders in the oil market. That is because even though the fall was large it took place in many small downward steps rather than large discontinuities (jumps) in the price.
The traders in the oil market who were long oil (set to gain from a rise and lose from a fall) had the opportunity to unwind their position as the oil price fell. That is, they could dynamically hedge their position to limit their losses.
At each point in time the ‘speculative traders’ in the market - the proprietory trading desks of investment banks, the hedge funds, the commodity trading firms, individual traders - take their positions. They position themselves to be compensated by hedgers for absorbing risk or they take a view on the direction of the market, or both.
Usually these speculative traders are highly leveraged; meaning that they take on a lot of risk relative to the capital that they have for absorbing losses. The price would not have to move much against them for their capital to be wiped out.
However, they are not passive. Instead, with every incremental movement in the market against them they reduce their exposure. As their exposure declines their losses are smaller and smaller with each additional adverse change in the price. The result is that no matter how much the market moves against them the sum of small losses never adds up to a fatal loss.
This process of adjusting after every small movement - decreasing exposure after adverse price movements and increasing exposure after beneficial movements - is variously known as ‘delta hedging’, ‘portfolio insurance’ or ‘dynamic hedging’.
Every trader dynamically hedges - they all take their positions and if the market moves against them they retreat quickly and live to fight another day.
What could possibly go wrong? The main risk in dynamic hedging is from discrete jumps in prices. Dynamic hedging relies on prices changing smoothly in small increments (deltas) and adjustment after each incremental change. If prices simply jump to a new level then a trader’s losses can be fatally large before the trader has a chance to reduce exposure. When a trader experiences a large loss from a discrete jump, they are said to be gamma’d (where gamma is the mathematical symbol for rate of change of delta). Traders are of course aware of the risk of being gamma’d and build it into their risk-return calculations.
Back to the SNB and the discrete jump in the SFr. The oil price fell by 50% over three months - no problem. Some traders booked losses, others booked profits, but they almost all survived due to dynamic hedging. Even if the price fell 50% over just a few days, instead of 3 months, it would be ok so long as each of the downward movements were small.
But on 15 January the SFr jumped 39% in 15 minutes and put the market into disarray. After about 40 minutes it settled at a gain of about 17%. Dynamic hedging could not operate properly through this process.
Based on the past experience of discrete jumps in high volume markets we should expect some very large losses from this episode. That was the case on Black Monday in October 1987 when the S&P500 jumped down 21%, and in the Asian banking crisis of 1997 and the Russian debt crisis of 1998 and several times already in the GFC.
The trading losses reported so far - $150 million by Deutsche and Citi, $100 million by Barclays, over $700 million by the Everest Global hedge fund - seem relatively low to me. I am expecting larger and more widespread losses to emerge, especially among hedge funds.
Brokers get gamma’d - so don’t leave cash with brokers
Brokers allow their clients to take risky positions whilst putting up only relatively small amounts of collateral (to absorb losses). A client might have short exposure (gaining from downward movement) to 1million SFr but only post 50,000 SFr in their account to absorb losses. The leverage in this example is 20-to-1, but brokers often allow much higher leverage.
If the market moves against the client (the SFr rises), then the broker takes the money out of the client’s account to recognize the loss and then insists on the client putting more money into their account. But what happens if the movement in the price is very large? If the SFr rises by more than 5% then the client’s account is in the red and if the client does not cover the loss, then the broker must absorb the loss.
Brokers are at risk of losses when big discrete changes in prices occur - brokers can get gamma’d. The more leverage they offer their clients the greater the danger. The jump in the SFr forced UK online broker Alpari into administration. The largest US retail FX broker, named FXCM, had to raise $300 million from outside sources to stay in business.
Sometimes retail clients of brokers leave their profits from trading in their brokerage account. That is a very bad practice. Unlike bank deposits, cash held in brokerage accounts is not guaranteed by the Federal Government. Brokers collapse from time to time and when they do some clients who have money with the broker suffer large losses. Don’t leave any more money with a broker than is strictly necessary. Sweep gains into a bank account every night and transfer money out of your bank account into a brokerage account only when strictly necessary.
Finally, if you find it shocking that a central bank of a country that is a global financial centre would behave in such a reckless and damaging manner, then I would say - get used to it. The failure of central banks to arrest the global slide into deflation is narrowing their options and forcing them into behaviour they would not have even considered just a few years ago. I will come to that next time.