LONDON (Reuters) - Italy’s recent bond market rout illustrates how popular risk- management models traders use can underestimate the impact of volatility in markets, according to research by a Bank of England official.
But gauges of market volatility across asset classes has slumped to near record lows, so models based on them can underestimate future market moves, Matt Roberts-Sklar, who works at the Bank of England’s capital markets division, wrote in a blog published on Tuesday on the “Bank Underground” website.
He noted that the models can be upended by a single-day shock such as the one seen during the May 29 selloff in Italian debt. That in turn can amplify selloffs by forcing traders to dump assets in a falling market.
Because these gauges use historical data on market movements at a time of low volatility, “the probability of a large move will be hugely underestimated,” Roberts-Sklar wrote.
He highlighted in particular “value-at-risk” models. Widely used in the financial industry, these use historical data of daily swings in financial assets to signal to traders whether they should add or cut back risk.
VAR models attempt to boil down the extent of risk a firm is taking across its trading strategies by quantifying the amount it may lose over a given time and attaching a probability to it.
Roberts-Sklar estimated the daily change in the “value at risk” in percentage terms nearly tripled on May 29 for two-year Italian government bonds to 170 percent.
“Part of the reason for the jump in the estimated VaR is the very low volatility in the prior few years,” he added.
On that day, the peak of Italian bond rout caused by a political crisis, two-year government bond yields soared over 150 basis points, surpassing even the 2009-2013 European debt crisis.
Latest European Central Bank figures show a record 33.7 billion of net foreign selling in Italian debt in May, according to Jefferies data.
Market shocks have become more common across financial markets in recent years, posing a challenge to investors who rely heavily on these models.
In the last three years, investors have had to face a sudden collapse in the British pound in October 2016, a surge in the Swiss franc in January 2015 and a jump in VIX earlier this year.
Reporting by Saikat Chatterjee; Graphic by Dhara Ranasinghe; additional reporting by David Milliken; editing by Sujata Rao, Larry King