Peer under the proverbial hood of a big bank or investment company's risk management chassis and one might see a mathematical model that looks something like this.
"Value at Risk," or VaR, models use statistical analysis and historical data to attempt to quantify how much an investor might expect to lose from trading within a certain time frame and within a certain probability. While such VaR models were criticized in the aftermath of the financial crisis for failing to predict heavy losses incurred on subprime mortgage securities and associated assets, they still form the backbone of Wall Street's risk management systems. Traders operate within their limits; woe betide the junior trader who exceeds their VaR.
But big market moves have been occurring more frequently in recent months, and that's wreaking havoc with VaR models built on certain expectations of the way markets fluctuate.
These "VaR shocks," as they're known, can be painful for investors because they often require them to cut positions in order to return to within their VaR limits.
Here is a terribly simplistic list we've drawn up, showing all the major VaR shocks we can remember in recent history.
- 2003, June : The prices of Japanese government bonds slumped suddenly, causing Japanese banks to bump up against their VaR limits and sell-off their JGB holdings.
- 2008, October: Widespread market turmoil causes a sharp uptick in VaR across the financial system.
- 2013, June: The "taper tantrum" sparks a selloff in U.S. Treasuries, with at least one bank reportedly breaching its VaR limit.
- 2014, Oct. 15: U.S. Treasury market suddenly "melts-up," causing investors to quickly reposition their portfolios.
- 2015, January: The Swiss National Bank unexpectedly removes its currency floor, causing a further VaR shock.
- 2015, May: Investors sell German government debt, with market participants labeling the event the latest VaR shock.
While these data points are largely subjective, it's not difficult to say that VaR shocks appear to be growing more frequent as big price changes haunt the market with increasing frequency.
Here's Bank of America Merrill Lynch credit strategist Barnaby Martin with a chart making that point. The number of assets registering large moves—four or more standard deviations away from their normal trading range—has been growing. Our back-of-the-envelope scribblings pin that kind of move as something that might be expected to happen every 62 years.
Why so many big market moves and so many VaR shocks? We've heard a few reasons.
For a start, the actions of central banks have suppressed volatility for years, enabling big banks and investors to assume more risk without bumping up against their internal VaR limits. Last year, the Treasury Borrowing Advisory Committee—an influential group of banks who advise the U.S. Treasury—warned that low volatility had created a "feedback loop" (PDF) that was affecting VaR models and causing many large investors to assume additional risk. When volatility does return to the market, it tends to wreak havoc on VaR models that have grown used to the steadiness induced by central banks. On a related note, since many VaR models run on five-year courses, the extreme volatility experienced during the financial crisis has been phased out of many VaR models. Fitch Ratings said last year that the trading VaR of the biggest U.S. banks had fallen by two-thirds from the end of 2010 to mid-2014, for instance.
To complicate matters, years of low interest rates have also herded investors into the same market positions; when they sudenly want to shed the positions, they all want to do it at the same time. Bond markets are also generally agreed to be less liquid, meaning that even meager trading volume can cause wild swings in the prices of certain assets.
Or as Jim Reid, strategist at Deutsche Bank, put it:
What has become increasingly clear over the last couple of years is that the combination of high money liquidity (Zirp and QE) and low trading liquidity (regulation and bank capital constraints) creates air pockets. The former encourages investors to move in a similar (positive) direction until overheating occurs with the latter then creating problems when they want to collectively lighten up. Those of us in the credit market saw this close up with [high-yield bonds] in [the second half of] last year. However that this is increasingly spreading up the top of the capital structure is a worry. It’s also a worry that these events are occurring in relatively upbeat markets. I can't help thinking that when the next downturn hits the lack of liquidity in various markets is going to be chaotic. These increasingly regular liquidity issues we're seeing might be a mild dress rehearsal.
Are these VaR shocks a bad thing? Yes, in that they can be very painful and their effects difficult to predict because they often affect other asset classes as investors reposition their portfolios.
But small VaR shocks can also be helpful in that they may help prepare investors for the inevitable coming of higher interest rates and—perhaps more important—potentially remind them of the pitfalls of over-reliance on backward-looking mathematical models in a world of unpredictable market events.