VaR – Value at Risk Model/s today are the bread and butter of FRM.
No application and /or discussion in the area of Investment Risk Management can develop or end without discussing value at risk modeling methods.
VaR Models are of different types and cater to different asset classes in different ways.
The three basic type of Asset- side VaR Models which are used to measure Financial Risk are:
- Market VaR Models
- Credit VaR Models
- Operational VaR Models
On the Liability side VaR Models have been developed and utilized by Pension Funds and Asset Management Companies to analyze Withdrawal Risk, Run on the Deposit Risk (Retail banks) and Funding (Asset -liability Gap) Risks.
So what distinguishes VaR from the other risk measures?
- VaR Models produce a summary statistic unlike other risk measures.
- VaR Models are a dedicated downside risk measure unlike other risk measures such as Altman Z Scores, Equity Beta, Bond Duration and Option Delta.
- VaR Models are a probabilistic forecast of loss/es unlike other risk measures.
- VaR Models are a graduated method of measuring risk.
- Market VaR Models only apply to mark to market exposures.
- VaR Models have raised the level of awareness with respect to “Market Risk Management” at Financial Institutions and has broadened the scope of Financial Risk Management.
- VaR Models have assisted Fund Managers and Bankers to develop risk haircuts on a more dynamic and forward looking basis.
- VaR Models with certain modifications can capture Non Linear and Higher Order Risks which other conventional risk measures cannot track!!
- VaR Models have assisted risk managers to develop more precise Stop Loss and Risk Limits.
- VaR Models have helped Stock and other exchanges to develop robust risk management, collateral management and margin pricing systems.
- VaR Models have enabled financial institutions to develop Economic Capital Models.
- VaR Models have given birth to a new level of reasoning and method of analysis within the existing discipline of Microeconomics (Financial Economics).
- VaR Models have introduced the concept of “Early Warning Signalling” across Asset Trading Markets => (however this may lead to building up of a higher systemic contagion risk factor)
Having said that, we recognize that VaR Models follow certain strict statistical and /or mathematical assumptions. Also such Risk Models are often subject to theoretical rigidities and limitations.
During the recent “Global Credit Crisis”, most of the Value at Risk type models were criticized for not measuring financial risk as required and hence investors were not able to hedge themselves on a pro-active basis.
The biggest drawback with such models is the assumption that the market will remain normal (within the forecasted volatility range as quantified by the standard deviation metric) and hence no crash or cataclysmic shock will be taking place during the trading hours.
How terribly unrealistic eh! ??
Nevertheless VaR models have not outlived their usefulness! Basel III Accord and its predecessor the “BASEL II” continues to rely on statistical measures such as VaR to develop the Economic Capital Models, CAR – Capital Adequacy Ratio, the related risk buffers and provisions, etc.
so has the time come for the financial industry to replace or keep VaR in its place?
That is a question that we and all the regulatory experts need to answer asap.
thank you for your precious time!
Your feedback is welcomed.
Kindly don’t confuse VaR value at risk models with Vector Auto-regression Models used in Econometrics!