Fed policy makes doing right by clients difficult

The following is excerpted from the quarterly commentary of Ben Inker, head of the Asset Allocation group for…

The following is excerpted from the quarterly commentary of Ben Inker, head of the Asset Allocation group for GMO LLC. To read the full commentary, click here.

Over the years, we at GMO have certainly done our share of Fed bashing. Most of our complaints have centered on the way in which overly accommodative monetary policy and a refusal to see the dangers of, or even the existence of, asset bubbles can lead to economic problems. We’re about to pile on the Fed again, but this time it’s personal. Our major complaint about Fed policy is not about the risks today’s ultra-loose monetary policy imposes on the global economy (which are considerable), but rather the fact that Fed policy makes it tricky for us to know whether we are doing the right thing on behalf of our clients.

One thing that we can say about the 2000 and 2007 asset bubbles is that, while they may have done significant damage to the economy and investors’ wealth, it was at least simple for us to know what to do with our portfolios. If we avoided the overvalued assets (which in 2007 was pretty much everything risky) we knew we were doing the right thing. Of course, even when investing is simple, it isn’t necessarily easy. In both episodes, but particularly 2000, the conservative portfolios we were running underperformed until the bubbles burst, causing plenty of consternation for our clients in the process.

Today, the Fed has engineered a situation in which the really unattractive asset classes are the ones we have always thought of as low risk: government bonds and cash. And unlike the internet and housing bubbles, this time it isn’t a quasi-inadvertent side effect of Fed policies, but a basic aim of them. The Fed has repeatedly said that a central part of the goal of low rates and quantitative easing is the creation of a wealth effect by pushing up the price of risky assets.
By keeping rates very low and taking government bonds out of circulation, the Fed is trying to entice investors into buying risky assets. The question we are grappling with today is whether we should take the bait.

So what makes this different from 2007? We’ve got some very unattractive assets and some others that look a good deal better by comparison. The trouble is that if those unattractive assets are cash and bonds today, moving to the relatively attractive assets involves increasing portfolio risk, whereas in 2007, moving away from risky assets lowered portfolio risk. In 2007, we could hold a portfolio that, whether assets took 7 years to revert to fair value or reverted tomorrow, we would still outperform. This was reassuring, because even though we use a 7-year reversion period in our forecasts, we know that the timing of mean reversion is highly uncertain. Today, the portfolio you would want to hold if assets were going to mean revert immediately is quite different from the one you would hold if you believed it would take 7 years to get back to fair value.

Perhaps the easiest way to think about the problem is to look at the efficient frontier for long-only absolute return portfolios, which we have reviewed on a number of occasions over the years.

By their nature, efficient frontiers are upward sloping with regard to risk. The major ways in which they change over time is the slope of the line and the level of the line. As investors, we all want the frontier to be high on the chart, which will occur when asset classes are generally priced to give strong returns. When the line is steep, it means you are getting paid a lot for taking on additional risk. Today’s line is very low, but reasonably steep. That means you are getting paid to take risk, but the reason is not very high returns to risky assets but very low returns to low risk assets.

Traditional quantitative analysis tends to assume that the level of the line is irrelevant. You might wish that the line was higher, but you can’t do anything about that. All you can do is decide how far out on the frontier you want to be, which is a function of your risk tolerance and the slope of the line. So, if we believe in our 7-year asset class forecasts, some reasonable forecast of volatility, and nothing else, we should be pushing out on the frontier, which is what Bernanke wants us to do.

In reality, the portfolios we are running for our clients are not particularly far out on the frontier. We are running at “normal” levels of risk or slightly lower than that. Why? You could chalk it up to a knee-jerk reluctance to do what a central banker is telling us, but when the asset allocation team is debating the appropriate level of risk to take in our portfolios, Bernanke’s name does not generally come up. The reason for our reticence to move out on the risk spectrum really comes from our idea of what drives portfolio risk. We believe strongly that the risk of an asset rises with its valuation. Stocks at fair value are less risky than stocks trading 30% above fair value because the expensive stocks give you the risk of loss associated with falling back to fair value.

That risk – “valuation risk” to use my colleague James Montier’s terminology – leads to losses that should not be expected to reverse themselves anytime soon. A cheap asset can certainly go down in price, but when it does, you should expect either high compound returns from there, which make your money back steadily, or a reasonably sharp recovery when the conditions that drove prices down dissipate, which will make your money back quickly. The loss is therefore temporary, although it may seem unpleasant while it is occurring. When an expensive asset falls back to fair value, subsequent returns should only be assumed to be normal, which means that the loss of wealth versus expectations is permanent.

Today, stocks are expensive relative to our estimate of long-term fair value. The trouble is, so are bonds and cash. If everything was guaranteed to revert to the mean over seven years, we would hold equity-heavy portfolios, because the gap between stocks and either bonds or cash is wider than normal. But we don’t know that it will take seven years. Because cash and (most) bonds have a shorter duration with regard to changes in their discount rate than stocks do, fast reversion would lead to smaller losses for them than for equities. Holding a portfolio where we are crossing our fingers that mean reversion will be slow is difficult to be excited about and, as a result, we are lighter on equities than the seven-year forecasts would otherwise suggest.

That leaves us around 63% to 64% in equities for a portfolio managed against a 65% equities/35% bonds benchmark and 48% to 58% in equities for absolute return oriented portfolios, depending on their aggressiveness and opportunity set. On the government bond side, given the incredibly low yields around, the only bonds we have much fondness for are Australian and New Zealand government bonds, because only those countries give a combination of a decent real yield and government spending policies that are sustainable in the long run. But our appetite for even these bonds is not great, leaving us with significant holdings of cash and “other.” If our seven-year forecasts play out exactly to plan, we are leaving some money on the table by not moving more heavily into stocks. However, our positioning does leave us in a place where we need not fear the circumstance whereby asset classes revert to fair value faster than expected, and that does help us sleep better at night.