Tuesday, November 14, 2017

Pension Actuaries: The Joke is On Us

...An Actuary is someone who wanted to be an accountant but didn't have the personality for it....An introverted actuary stares at his own feet, and extroverted one stares at the other person's feet....What is the difference between God and an actuary?  God doesn't think he is an actuary...."Look at the white horses over there." Actuary: "They're white on this side, anyway."

Unfortunately, when it comes to the mess our pensions are in, actuaries are no joke. Pension funds labor under actuarial assumption for expected returns that are mainly pulled out of thin air without any regard for financial economics. Does anyone really think pensions will be able to grow at seven percent or more per annum? When they are constrained by their various constituents to hold 50% to 60% in bonds and cash? (See Figure 3 of this OECD publication.)

Those unrealistic assumptions lead to unsupportable levels of contributions, and thus pensions are underfunded as a matter of course. You know the assumption are wrong when virtually every pension is on the negative side of the ledger.  But the actuaries do not seem to have been trained in the concept of making course corrections. Nor do they educate themselves in financial economics to better evaluate the mess they are creating.

Why am I going into this? Over the weekend I attended an event honoring a former colleague of mine from my days at Morgan Stanley, Jeremy Gold.  He is an actuary who has spent his career trying to move the pension actuaries toward a firmer foundation in financial economics -- to have them at least avail themselves of what financial economists can provide. In the mid-1980's, Jeremy and I worked together in the Fixed Income Research Group at Morgan Stanley. He and I went on various trips to push fixed income products, and to market the fledging new strategy of portfolio insurance (flying out once to have dinner with the head of the Port Authority of Los Angeles, who we discovered, part way through the first course, had -- surprise -- thought we were selling port insurance). We co-authored a paper in 1988, In Search of the Liability Asset, that is still on various reading lists, maybe not for actuaries, but at least for those in finance.

Jeremy left Morgan Stanley in the late 1980's to get a Ph.D. from Wharton, and spent the next twenty years as a thorn in the side of the actuarial profession, pushing them to add financial and economic structure to their methods. One of the best and most widely read of his works for this is the paper Reinventing Pension Actuarial Science.

At the base of it, finance is not actuarial science.  It is not predicated on repeatable, or even known, probabilities. There is no appeal to the law of large numbers for the systematic risks of the financial system.The future does not look like the past. There are no mortality tables for asset returns.

...What do actuaries and Packer fans have in common? They both think that history will repeat itself...

So if you want to be the arbiter for over 20 trillion of U.S. pension assets, a good start is to do it based upon a foundation in finance.


Monday, November 13, 2017

More on ETFs -- A Little Craziness in High Yield Bond ETFs

I wrote a post a month or so ago on the risks from ETFs, in particular how ETFs on less liquid markets -- with high yield bonds being my poster child -- could cause problems for the market generally.  Basically that there is a fundamental flaw when people think an instrument based on a illiquid market is capable of intraday liquidity. And that if the ETFs in such a market have a severe problem, ETFs generally might be considered tainted by a range of retail investors, leading to an outflow from even the more liquid ETFs.

Here is an article that points toward the potential for problems with high yield bond ETFs:
Investors Playing ETF Rout Pushed Junk Bonds to Brink of Chaos. 

Excerpt:

Trading in exchange-traded funds got a little crazy last week when it became clear that junk bonds were in for more pain. But the market was fortunate the consequences weren’t more severe, strategists warn. Though spared the worst, investors came close to creating a scenario where ETF activity drove prices....a snowball effect where a dislocation develops between the fund price and the value of its underlying assets.

Saturday, November 4, 2017

Our Low Risk (Low Volatility) World

In case you haven't noticed -- and I haven't -- we apparently are in a world of exceptionally low risk. To see this you need look no further than the volatility of the major markets. The volatility of the U.S. equity market, for one, is at its lowest level in a generation. So, no worries here, right?

I wrote a blog post in 2011 titled The Volatility Paradox which explained that when volatility is low, risk is actually rising because people are more emboldened to take on higher leverage and to move to riskier assets. If volatility is half of what it used to be, why not lever twice as much? Thus the immediate question is what happens if there is a sudden surge in volatility from our current, low level. What is the dynamic through which a volatility shock might propagate across the financial system?

A conventional stress test will assess positions that have explicit volatility exposure, such as positions in options, in the VIX and other volatility-based instruments. There is plenty of dry powder here; over the course of 2017, as the U.S. equity market volatility as measured by the VIX index dropped to one of its lowest points in history, we have seen a growing concentration in short volatility exposure by leveraged ETFs, mutual funds, and hedge funds.

But a stress test that does the simple mathematical calculation of direct portfolio exposure to volatility will underestimate the effect of a rise in volatility, because there are dynamics triggered by other strategies that do not have explicit volatility exposure but that have a link to the volatility of assets and to the assets themselves. A rise in volatility will trigger actions for these strategies, leading to selling of the underlying assets, and this in turn will lead volatility to rise even more, creating a positive feedback between the volatility of the market and the assets in the market.

What are these strategies? Well, a good place to start are volatility targeting, risk parity, and other strategies that will rebalance their portfolios when volatility rises. Volatility targeting is a strategy that targets a level of volatility to manage risk that is typically set based on the manager’s mandate. For example, the manager might follow a strategy that will seek to keep the portfolio’s volatility near 12%. If the volatility of the market is 12%, the fund can be fully invested. However, if the volatility of the market rises to 24%, the fund will sell half of its holdings in order to stay in line with its 12% target. Risk parity allocates portfolio weights to have the same total dollar volatility in each asset class. These multi-asset class strategies often use leverage to adjust holdings of underlying assets, buying more of the lower volatility assets relative to the higher volatility ones. If the volatility of one of the asset classes rises, the fund will need to sell some of that asset class in order to maintain equal dollar volatility. In both of these cases, there is a clear mechanism going from the rise in volatility to a drop in the underlying asset.

Things won't all happen at once. The agents for these strategies differ significantly in their time horizon. Those who are directly linked to volatility, those that are in short-volatility ETFs and the like, will have an immediate P&L effect from a surge in volatility, and will need to reduce their positions immediately. The volatility targeters will only reduce positions as the rise in volatility is seen as having a non-transient component, and their adjustments will be in a weekly to monthly time scale.  And the risk parity agents will have an even longer time horizon, because they generally make asset adjustment with a monthly or quarterly time frame, and do readjustments based on a longer-term estimate of volatility.

Oh, a little more on the relationship between low volatility and risk: If you want to see really low volatility, look back at the swaps markets in the summer of 1998. But you might recall that in August of 1998 there was a rash of defaults in Russia, which was then followed by the blow up of LTCM, the (before that) famously successful quant fund whose principals a few months earlier had graced the cover of Business Week (never a good sign). In that case, low volatility didn't spell low perceptions of risk, it was an indication that no one wanted to go into those markets because things were so uncertain. I go through my first-hand experiences with this in one of the chapters of my 2007 book, A Demon of Our Own Design.

Thursday, November 2, 2017

Interview in the U.K. for the Daily Mail

I had a video interview in the U.K. on my recent book, The End of Theory, with Rachel Straus, (Big Money Questions), that just came out with the Daily Mail. You can get to it here.