Saturday, November 22, 2014

Value at risk Shock

On october 15th, There was a turmoil in the US treasury security market. The yield on the benchmark 10 year US government bond dropped by 33 basis point to 1.86% before settling to 2.13%. This does not look much move but market participant confirmed that this was seven standard deviation move from the intraday average. Such event can occurs only once in 1.5bn years.

After this huge market volatility analyst tried to figure out the reason and even trader and regulators also wanted a understanding.
One of the explanation is provided as VaR shock. After the financial crisis, major dealer banks have curbed the capital allocation to the market making activity and VaR limits have imposed to trader or desk. If VaR limit gets breached than trader or desk has to sell off these positions. With the shockwaves of late 2008 now gradually receding, and a period of low market volatility taking its place, these VaR models have been indicating that the risk of investors sustaining large losses is very low.That means investors may be subject to a so-called “VaR shock” in the event that volatility returns to markets.
This is what happened in the market as yields dropped in the market and computer algorithms used by market making systems start buying more treasuries in order to stem their losses.
Lower risk appetite at the big dealer banks have created the effect of reducing liquidity in trading security. Market liquidity get worse in selloff periods and dealers positions, long or short, declines in the sell off periods. Dealers reduction in net positions is associated with the reduction in dealer's risk,VaR. 

“VaR-based analysis leads to self-reinforcing loops,” a group of banks warned in a presentation to the US Treasury weeks before October 15. “An unexpected increase in volatility might come from broad-based selling of assets wanting to de-risk in front of a turn of policy.”