We've entered a new dawn of higher volatility21 января 2015, 18:18
For better or worse, I have been trading FX for more than 30 years.
Coincidently, I started trading around the time my native Denmark’s currency, the DKK, was pegged to the D-Mark in 1982/1983.
• Discontinuous pricing is the biggest risk of any market.
• Lack of liquidity has and will always be the single biggest risk for any position held in the market.
• The mathematical models of markets are flawed due to underlying assumptions (for example, VaR and risk models generally)
• Capital requirements on banks (market makers) were already increasing before the SNB move as a longer term result from the global financial crisis and need for lower total leverage
• There is growing national regulation on maximum risk-taking.
What is volatility?
A text book definition: A measure of variation of price of a financial instrument over time. The realized volatility is how much volatility actually happens over time, while the market pricing of potential future volatility is called the implied volatility and is derived from how the market prices derivatives. The expected volatility.
Options pricing: the fat tail/skew risk/volatility “smile”
The tail-risk in options pricing is the most difficult. You all know our Outrageous Prediction publication, or what some wrongly call: Saxo Bank’s Black Swan events. That’s not a fair label, because option theory has operated with a volatility skew or “smile” since well before Nassim Taleb coined a better marketing phrase (the Black Swan) for the same phenomenon. In other words, it is more expensive relative to the perceived probability to buy insurance on unlikely events – hence the “smile”, or higher price in options at strike prices far above and below the market price.
The irony is that, while there is a perceived fat tail risk and therefore a “smile” in options pricing, during traumatic market events, the pricing sometime proves far too cheap. So the statistical risk and hence the market risk of options lies in the pricing of “tiny delta”, or supposedly very low probability events, because Black-Scholes and newer options pricing models do not correctly price these unlikely events – like CHF moving 35% as we saw last week. In other words, both in terms of statistical probabilities/options models and even for veteran observers like me of 30 years of real market histories covering some spectacular market events, this CHF move was bigger and faster than anything I have ever seen. This event was the fattest of fat tail risks – an event that was perceived as extremely unlikely, and that very perception was what allowed the leverage to develop, via mis-priced derivatives, that in turn aggravated the move when it unfolded. As well, the situation was made even worse by no hand in the market large enough to step into the market to soften the blow. It was, after all, the SNB itself that stepped away from its own policy.
Discontinuous pricing / price gaps
Discontinuous pricing is extremely important because during traumatic situations, we not only have to deal with “flawed mathematical models” that plague options pricing (which work the vast majority of the time, but don’t do a good job of predicting intense market events) but we also have to deal with flawed assumptions about how much the market can move during a fat-tail event. By far the worst risk for a position during these events is the assumption that prices are continuous. For example, there is the assumption that if EURUSD trades 05/07 on the bid/offer, then the next “expected” trade level is 06/08 or 04/06 as the market moves up or down, with a possible widening of the bid/ask or a fast move of 20-30 pips on the release of key economic data (Like US nonfarm payrolls).
But in times of extreme distress, like when the SNB removed the EUR/CHF floor last Thursday, the market was virtually unable to transact in any size from 1.2000 bid to .8500 offered because in an instant there were almost only sellers and practically no buyers. In normal trading conditions there will up to tens of thousands of quotes per second in a major currency pair, but last Thursday, the ability to transact CHF electronically or via market making virtually disappeared altogether for almost an hour – in other words because sufficient EURCHF buyers could not be found to counter the avalanche of those wanting to sell – the market can only execute a price if there is a price, hence the trading regime was one of discontinuous pricing.
Even under normal market conditions, it’s more than common to see huge “air pockets” of missing liquidity. From 1997/2000 I ran trading in New York for a bank and later joined the prop desk at UBS. There, I could sometime trade 100 million EUR without even moving the spread and sometimes the market would move 50-60 pips as the order was executed.
That brings us to liquidity.
Illiquidity: The biggest single risk, along with correlations moving to 1
The biggest risk for any trader: private, bank, hedge fund or small retail client is discontinuous pricing (essentially, when the market disappears entirely for gaps along the price continuum) and illiquidity, which can become a systemic problem. Here’s why: the response from regulators to avoid market players taking on too much market risk has been to introduce VaR – or Value at Risk – requiring that your positions are able to survive moves that are demonstrably not likely to happen 95% of the time. This is fine as an academic exercise but remember the fat tail we discussed above and then realize the self-fulfilling problem that occurs due to this VaR rule when markets go “crazy”:
1.) All assets classes correlations move to one – i.e: This happens because during a crazy market event, there is a move to “safety” with moves in major assets increasingly moving in synch. This is often bonds, US dollar and Gold all rising and equities falling, for example. The synchronous moves across markets causes a self-reinforcing explosion in VaR calculations, which forces portfolio managers/hedge funds to hedge in the same direction as everyone else who is also buying “safety”.
2.) Due to the sheer volume of trades in the same direction prices “gap” or become discontinuous.
Momentum drives markets
So what happens with risk control and trading conditions during “force majeur” or major moves (larger than 3 standard deviations) is that all models on risk explode and create more momentum to safety assets.
What does this mean going forward?
I have no doubt that margin requirements for traders will increase from here (requiring a reduction in leverage/gearing). The SNB made crystal clear to all of us that clients, funds and banks did not have “bad enough” risk scenarios embedded in their risk controls, and the response will inevitably be higher margin and possibly increased spreads down the line.
FED and US regulators already doing margin calls on banks’ capital – Maximum leverage now down to 20 times
The banking system overall is meeting stricter and stricter criteria on capital set aside for events like this one – actually a fairly rational move considering that VaR does not fully cover “worst case scenarios”. We already saw the US rolling out stricter capital demand requirements on US banks in 2014: Big US banks must boost capital by $68 billion under new rules.
‘The rules require the eight biggest bank holding companies to maintain top-tier capital equal to 5 percent of total assets. Insured bank subsidiaries must meet a 6 percent ratio. That's higher than the 3 percent ratio included in the Basel agreement’ (Source: Reuters)
This simply means that the largest US banks increasingly need “real capital” to do business and the maximum leverage they have is 20x (5% margin). This includes the whole balance sheet. Meanwhile, the BIS allows 3% or 33.3x leverage. I am absolutely sure the US standard will be global standard before 2020. If the banks, the world’s main market makers, are under capital restraints, the FX market overall will see significant lower supply of high leverage accounts. I believe, personally, that Singapore’s maximum leverage of 5% for FX and the US’s of 3% will become international norms. The pressure is on to reduce systemic risk and it makes even more sense if my piece from yesterday is correct in assuming a paradigm shift: End game for central banks.
The pretend-and-extend model is running out of time
The last eight years of economic ‘experiments’ is unusual in many ways: We have had the single biggest intervention in the economy, through central banks and their different shapes of monetary stimulus, in history.
The central bank’s interventionism suspended the normal business cycle, a cycle that in normal times cleans out poorly run businesses and allows only the stronger to survive in a constant cycle of decline and renewal. There has, in my opinion, been an “artificial equilibrium” in place since the March 2009 low in the S&P 500 index at 666, as central banks have aimed, with good, if misguided intentions, to buy more time and see if the economy can heal. A derivative of their efforts has been artificially low interest rates that have mispriced money and artificially low volatility that has mispriced just about everything. Denying the business cycle has had highs costs indeed.
Good intentions are rarely enough. The end of 2014 saw a 50% drop in energy prices. Oil is one of the few “freely” traded asset classes left and through the energy collapse “Mr. Market” just sent us a warning: the ruble has collapsed and even a sometimes safe haven currency, the Norwegian krone devalued as much as 15% in the space of a couple of weeks. Were these the advanced warnings of the SNB’s need to remove the EURCHF floor? With the convenience of perfect hindsight, I suspect they were. And this is what it may mean: Central banks are running out of time
The bottom line is: The SNB ran out of time, the ECB will run out of time this week, and the Fed, Bank of Japan and the Bank of England ran out of time in 2014. What comes now is a new reality – the SNB move was a true paradigm shift – we can no longer look at central banks, the markets and extend-and-pretend in the same light as we did last Wednesday (the day before the SNB pounced). The market price for this new paradigm shift will be higher volatility, lower leverage, and the risk of discontinuous pricing and poor liquidity, especially as central banks, pull away from the false equilibrium they have provided in the years since the financial crisis.
Chief Economist Steen Jakobsen