Asserting superior performance of its "standard" fat-tailed risk measures and post-modern methodology during the continuing credit and liquidity crisis, FinAnalytica Inc., leading provider of "glass box" risk management and portfolio construction analytics, today released complimentary access to daily fat-tailed VaR and Expected Tail Loss (ETL) estimates for major broad equity markets. Supporting these estimates, an Executive Briefing series, aimed initially at describing the limitations of conventional VaR estimates during this period, has also been launched.
FinAnalytica’s Cognity, the risk management industry’s only platform with integrated fat-tailed, skewed VaR and ETL distribution models across the entire risk measurement process, transparently calculates and backtests fat-tailed VaR and fat-tailed ETL estimates across global indices. It compares them to the classical Gaussian models employed by competitive systems. The results are irrefutable: fat-tailed risk estimates consistently outperform normal distribution based risk assumptions in both the current crisis and over longer historical time windows.
Zari Rachev, chief scientist of FinAnalytica, explained, "FinAnalytica is steeped in the ethos that risk must be managed by modeling fat-tailed and skewed returns and accurately estimating tail dependencies. Unlike many approaches which presume ‘crisis’ events are limited to ‘thousand year’ intervals, our methods are sensitive to and rapidly reflect extreme market turbulence as a matter of course, obviating any need to introduce proprietary methods ‘after the event’ in order to explain market shocks."
The daily fat-tailed VaR estimates are provided initially for twelve global indices including Dow Jones Industrial, NASDAQ, S&P 500, Russell 2000 and selected global MSCI indices.
The skewed fat-tailed models, combined with sophisticated volatility clustering techniques, are much more reactive to crisis conditions. This is clearly evidenced by the out-of-sample backtest results showing exceedances well within the stated confidence interval.
Analysis across all broad equity markets shows the advantages of using fat-tailed VaR both during crisis periods as well as in less volatile times. Specifically, results for the Standard & Poor’s 500 using a 99% Normal VaR led to 17 instances of realized loss exceeding the predicted VaR threshold over a 400 day window ending November 14, 2008. This is significantly in excess of the maximum 7 exceedances expected with 95% confidence. Conversely, using the fat-tailed VaR resulted in 5 observed exceedances of the VaR threshold over the same period, well within the confidence bounds and vastly superior to the Normal VaR estimates.
"Fund managers and risk officers globally are realizing the shortcomings of conventional volatility-based risk assumptions," said Doug Martin, chairman and ceo of FinAnalytica. "Providing these daily estimates along with the Executive Briefings will highlight best practices for accurately measuring risk and proactively rebalancing portfolios from a real world view."
Focusing initially on the current market crisis, the Executive Briefing series aims to demystify important concepts such as risk asymmetry, fat-tails and dynamic correlation shifts. The first briefing, "The Truth about Fat-Tails and Black Swans" was released in October and the latest edition, "Rebuilding Trust between Managers and Investors" is released today.
Marc Gross, managing director, added, "It is critical for investment professionals to understand that advanced risk management and portfolio construction tools can play a vital role in growing their business by objectively demonstrating added value and fostering engaging and productive dialogue with investors. Tools such as tail-risk budgeting can also help support incentive structures which decentralize risk management decisions by accounting for the tail-risk employed in achieving return."
