- April 22, 2015
For Myles Allen, Professor of Geosystem Science at Oxford University’s School of Geography and the ... [more]
Just when you thought you were safe, the next reiteration of risk parity has arrived. AllianceBernstein’s tail risk parity takes the concept of risk parity, reallocating assets uniformly according to risk, but it uses tail risk, not volatility, as the core measure.
The concept of risk parity is a portfolio diversified according to risk, rather than capital allocations. Traditionally, risk parity has used sources of volatility as the risk measure, but now it is being argued that this is not very helpful in times of stress. The newest iteration of the risk management technology is tail risk parity, which enables diversification of the sources of tail risk.
“If you believe that all markets are normal and there are no risk tails in any asset classes, then risk parity will work. But by their nature asset classes do exhibit fat tails, so because of that volatility doesn’t work,” says Michael DePalma, New York-based chief investment officer of quantitative investment strategies at AllianceBernstein.
“With risk parity you are ‘smooshing’ the tails in. It gives you a smoother ride, but it reduces the left and right tail. The cost of risk parity is you give up the upside,” he says. “In tail risk parity, by focusing on the left tail risk as a measure of risk, you can retain proportionally more of the upside.”
In this way, he says, tail risk parity is a positive skew, as it is reducing the skew to the left tail risk.
Alliance Bernstein is saying, while this is a unique approach to a problem, it is also the natural evolution of thinking around diversification benefits and protection.
DePalma says the popularity in risk parity strategies is due to the ability to solve a number of the shortcomings of a static rebalanced portfolio such as concentrated risk exposures and they maintain a constant risk target over time.
“Maintaining a risk target over time means you can harvest returns when you’re best getting paid for it,” he says. “But there are problems that risk parity hasn’t solved such as diversification failing when normal correlations are used to structure portfolios.”
Conventional risk parity strategies use volatility as the measure of risk, this works when markets are “normal”, DePalma says, but when markets are under stress then volatility doesn’t help you at all.
“When equity markets are exhibiting left tail risk, then 80 to 90 per cent of the time commodities are in their tail and credits and infrastructure will also have a bad return,” he says. “But in the middle 80 per cent, or normal times, it is 50:50 whether other asset classes do well or not.”
The AllianceBernstein strategy, which was developed with contributions by Myron Scholes, Nobel Laureate and co-creator of the Black-Scholes option-pricing model and currently the Frank E Buck professor of finance, emeritus at Stanford University, computes the tail risk for each asset class, which is implied from the options market.
“Using the collective wisdom of the options market, rather than building a measure, we come up with the market-implied level of tail risk,” DePalma says.
The strategy takes the universe of assets to include in the portfolio and assigns each asset class to one of three risk buckets – growth, safety and inflation – to ensure macro diversification. It then imputes the expected tail loss or the computed tail risk from the options market, and forms parity across the buckets with each contributing one third of the tail risk of the portfolio.
The tail risk level that is set is maintained using leverage in the form of futures and swaps.
The process, including the options market-imputed expected tail losses and tail risk allocations, is reviewed daily, but the manager doesn’t necessarily trade every day.
While most investors have looked at risk parity allocations as part of their alternatives allocations, DePalma says there is also an option to include the tail risk parity strategy as an overlay.
“At its core, this is a risk measurement and management technology. We can use futures and swaps across the portfolio and re-allocate the tail risk from the existing assets to create a balance,” he says.
The idea of downside protection is not new, but the strategy combines it with the existing advancements in asset allocation strategy that risk parity provides.
In addition, it claims to be more cost effective than buying protection directly in the options market. (This is laid out in the paper by Scholes, DePalma and AllianceBerntsein’s Ashwin Alankar, An introduction to tail risk parity: balancing risk to achieve downside protection).
But while DePalma says tail risk parity is a unique approach to solving a problem, it doesn’t solve all the problems of risk parity.
For one, the approach takes a certain measure of tail risk, namely the options market. So it won’t protect a portfolio against a flash crash or an earthquake.
However, DePalma says the largest risk is it may make the portfolio overly conservative.
“The biggest risk is the opportunity cost. I’d love to find a way to reduce that,” he says.
“This is a unique approach to solving a problem, and it may get the competitive juices stimulated and investors can expect more innovation,” he says. “In risk-managed solutions everyone is taking a slightly different approach and it behoves investors to look at a suite of offerings.”