Recent Posts

Wednesday, July 20, 2016

Bond Active Management -- Makes As Much Sense As It Ever Did

Cullen Roche (Pragmatic Capitalism; Orcam Asset Management) wrote an interesting article on the secular bond bull market, and the somewhat dismal forward-looking nature of bond returns. I find it hard to disagree with the bulk of the article, but I disagree with his sentiment about bond fund active management. In my view, the case for active bond fund management has not really been changed by the drop in yields. (Note that this assessment is deliberately ambiguous; I am not saying that the case for active management was good when yields were higher; I leave that assessment to the reader.)

The passage that caught my eye (and which is what I objected to):
This is particularly interesting within the context of actively managed bond funds. The fixed income space is often cited as a place where active management can actually make a lot of sense because there is the potential for greater strategic diversification than we generally see in the stock market. But it’s becoming increasingly difficult to squeeze alpha out of the fixed income stone in a falling interest rate environment. 
I will immediately note that I am approaching this question from an institutional perspective that is different from Cullen Roche's, and certainly different than most retail investors'. (As a disclaimer, my own fixed income portfolio consists only of bonds that I am holding to maturity, and fixed income exchange-traded funds; that is, no actively managed bond funds. My attitude is that the only person who is going to gamble with my bond allocation is me.)

Please note that what is discussed here is not the size of your bond allocation (for which absolute returns matter), rather whether your allocation should be actively or passively managed.

By way of background, I held a number of analyst positions at my previous employer. We managed a number of fixed income funds (large by Canadian standards) for a set of clients. The funds included actively managed, as well as index funds.

All Funds Are Active, The Question Is: How Active?

For a fund manager, you measure risk relative to your benchmark. (Your client is the one who has to worry whether the benchmark matches their investment needs.) You are constrained in how much "risk" you can take against that index, which will assume is measured by the Value at Risk (VaR).

(I know that there a lot of complaints about VaR. It is clear that it may have problems for measuring the true "risk" of a portfolio. At the same time, we need a quantitative measure that allows investment committees to allocate risks between different managers. If we rely on fuzzy assessments by risk officers, all the risk ends up being allocated to the personal favourites of the risk officers.)

I am unaware of any fund that has zero risk (VaR) versus a public benchmark. (You could invent a private benchmark that matches your chosen portfolio, but you would have tracking error versus the public indices that everyone else uses.) Other than for some very simple indices, it would be crazy (and extremely expensive) to rebalance your portfolio in response to every index event.

(The most popular Canadian indices rebalance daily; which means that the index weights change every day that the there is new issuance, a bond dropping below one year maturity, or a coupon payment. In other words, almost every single working day for the broad index.)

An index fund is going to do is follow rules that keep VaR "small," and attempt to transact in a fashion that keeps bid/offer drag minimal. The exact same thing is true for almost all fixed income funds. The only difference is that "active funds" will have enough of a VaR risk budget so that it can take positions that at least cover management fees, if not outperform the index.

A hedge fund is just an exaggerated version of this. They have larger management fees, so they need to take larger active risks (relative to assets under management) in order to have a chance of meeting their return target. (Since the target returns are so high, they do not use a bond index as their benchmark.)

If you invest in a bond index fund, you should expect to underperform the index by at the amount of the management fee, with almost no chance of beating the index. The "active" funds that I had the most contact with were only trying to beat the index by 20 basis points after fees (which were around 10 basis points). Even in a low return environment, it is hard to argue that 30 basis point annual outperformance (before fees) is impossible.

Presumably, Cullen Roche has a more "active" version of active management in mind (attempting to beat the index by 100+ basis points). Can such higher return targets be met?

Hitting Higher Alpha Targets - Why Not?

Equity-oriented investors tend to look at absolute returns and yields, and not realise that fixed income investors can always achieve "equity-like" 20% returns; all they need to do is crank up the risk dial. (Please note that this is probably a bad idea.)

This is achieved by:
  • Currency exposure. (Normally, there is a "separation of church and state" principle which allocates currency risk to a currency team, but there is nothing stopping a firm from leaving the two teams within a single fund.) My U.S. Treasury ETF's crushed my domestic Canadian fixed income benchmarks when the Canadian dollar collapsed below par.
  • Leverage and/or derivatives. If expected volatility is lower, you just dial up the size of your notional positions so that you keep your expected risk/return unchanged. (One of the most exciting topics in fixed income quant-land is the relationship between the level of yields and yield volatility; we tend to see that lower yields breeds lower volatility.)
To be clear, I am not saying that cranking up leverage is a good idea. In fact, if you are a retail investor, it is a very bad idea. (Unless you have a trading desk on call at all hours every working day, how exactly do you expect to manage your positions?)

However, it is probably a bad idea for institutional investors as well. As Hyman Minsky observed, stability is destabilising. As investors increase their leverage, they artificially depress volatility -- until they all have to unwind positions. This is a recurring theme in "hedge fund space."

It should be noted that these blowups are not just associated with low yields. Fixed income quant funds created a pretty impressive financial crisis back in 1998 (in my first few months in finance).

However, extreme leverage is only needed if you want to hit 20% returns. If you have a reasonable outperformance objective (50-100 basis points, say), it could be achieved by long-only investing. You can overweight corporate credits, take currency exposure, relative value (curve positioning), and possibly market-timing. Only the latter two are affected by lower yield volatility, and this can be compensated for by increasing position sizes by reasonable amounts (for example, instead of being short index duration by "six months," make it "one year").

Yes, it is unclear whether any particular fixed income fund will achieve this. At the same, it is unclear whether actively managed equity funds will do any better in practice. Equities feature greater return dispersion, allowing a greater potential tracking error for a long-only portfolio; but you need to find a manager that can take advantage of that tracking error systematically.

(c) Brian Romanchuk 2016

No comments:

Post a Comment

Note: Posts are manually moderated, with a varying delay. Some disappear.

The comment section here is largely dead. My Substack or Twitter are better places to have a conversation.

Given that this is largely a backup way to reach me, I am going to reject posts that annoy me. Please post lengthy essays elsewhere.