Tuesday, March 24th, 2015

One Way Street

We recommend that our clients use ETFs to execute some of the investment ideas generated by our research. We are therefore conflicted by what we are about to argue. In recent weeks there has been much comment on the hunt for yield and the possible overvaluation of investment grade and high yield bonds. We are not going to comment on valuation, because everything that was said about a normal level before QE has been massively disproved by events.

We are however happy to comment about volatility which has recently been towards the bottom end of its range for the last 20 years. We think this is important, because for any given level of valuation, investors will hold more of an asset if the risks attached to it are (a) lower than normal or (b) lower than the available alternatives. There is no secret about the reason for this low volatility. The principal way in which QE works is not by lowering interest rates, but by lowering volatility, which allows capital market intermediaries to hold more inventory for any given level of capital reserves. This applies even if they are subsequently told by the regulator to boost their capital reserves.

Similar logic also applies to portfolio investors. Any firm which uses a standard optimisation process will have come up against an interesting dilemma. If you take the raw data for total returns and volatility for the last three years in US dollars, no matter how many types of equity or alternative assets you input into the model, you only need three asset classes to optimise your return per unit of risk. They are large cap US equities, US investment grade debt, and medium-dated US treasuries. The proportions will vary depending on the objective function, but the portfolio will typically have an allocation of 35% to investment grade corporate bonds. It will not own any alternative assets like commodities, any non-US equities or any other fixed income apart from treasuries.

Most institutional investors are anxious to avoid this level of concentration. So they input their own estimate of future returns, and come up with a much more diversified portfolio. Nonetheless the raw data do have an impact and it is clear that many types of investor now have a greater exposure to corporate bonds than they did five years ago. This includes retail investors who have used the development of the ETF industry to gain significant exposure to this asset class for the first time.

What concerns us is what happens when the Fed start to raise interest rates. It seems to us inevitable that the volatility of fixed income returns will rise – at least to the median of the last 20 years and possibly higher. We already know, because they have told us, that the major market-makers have much less inventory-holding capacity than previously. Most portfolio optimisation processes will automatically recommend a reduction in exposure to corporate bonds even if yields appear more attractive. Retail investors will notice the underperformance of their “safe” ETF investment and sell. At the height of the taper tantrum, LQD the largest investment grade ETF, saw a peak daily outflow of $870m. This equates to 4% of the current market capitalisation of the fund and was a higher proportion then, and the selling pressure went on for weeks not days. If we get a repeat performance later this year, we are not sure that the infrastructure of the corporate bond market or the retail ETF industry can cope.

Leave a Reply

Full Blog Archive:

Other Recent Posts