PsyFi Search

Tuesday 28 April 2009

Arnott: 40 Years of Bonds Beating Equities

Bonds: Why Bother?

Rob Arnott’s paper Bonds: Why Bother? sets out to succinctly punch a few holes in accepted market norms. What he finds should be sobering for investors in shares who believe that they can gain excess returns for taking the excess risk of holding equities over reasonable lengths of time.

Now a lot depends on your definition of a “reasonable length of time” but as the paper shows that someone holding and rolling over 20 year Treasury bonds would have beaten an investor in the S&P 500 over any time frame since 1979 you'd have to have a very long term view indeed to regard equity investment as a safe thing. In fact bonds have actually beaten equities on the same measure over the last forty years.

Lower Equity-Risk Premiums Mean Equities Are Higher Risk

The contention here is fairly simple, that the equity risk premium of equities over bonds – the expected excess returns of shares – is much lower than commonly assumed. Arnott and Bernstein in What Risk Premium Is Normal? estimated it at closer to 2.5% than the generally anticipated 5%. That difference will compound up into a huge excess return over very long periods. However, it’s also sufficiently small that the excess risk of equities can wipe out the gains over shorter time frames. Unfortunately it seems that those shorter time frames can amount to up to an average person's entire investing lifetime.

So from 1803 to 1871, from 1929 to 1949 and now from 1968 to 2009 bonds have outperformed stocks. Of course, in all cases these will be comparing peaks to troughs – the stock market in 1969 was riding high and now in 2009 it’s on its uppers. This, though, is irrelevant for individual investors who have to live in the times they’re presented with: such levels of stockmarket underperformance can destroy a lifetime’s savings. You don't want to be retiring now.

Reading between the lines it looks like the stockmarket returns quoted are at the level of the index and don’t include the effect of dividends. What this means – as if we needed to be told – is that dividends and their reinvestment are critical to making any money out of the markets, once you adjust indices for inflation. When, as has happened over the past few decades, dividend yields drop to lows even those potential returns are endangered.

Deep Thought

The situation reminds me of the story in Douglas Adams' Hitchhikers Guide To The Galaxy where a group of philosophers constructed a computer called Deep Thought to find the "ultimate answer to Life, the Universe and Everything". The computer whirred away for a seven and a half million years before proudly announcing that the answer was "Forty-Two". Of course, no one had thought to ask what the question actually was.

In our case the answer we're given is "buy shares" but once again no one seems to have wondered about what question we're trying to answer. If it's "what asset class gives the best returns over indefinite timescales" then that's probably about right. However, if it's the rather more useful "what should I buy to ensure a safe retirement" then we probably need to think again.

Indexing and Returns

Rob Arnott is one of the thought leaders in the world of fundamental indexing – an approach to index tracking that believes that indexing on market capitalisation, as most indexes do, is bound to overweight expensive shares to the detriment of investor returns. However, the new paper focuses on equity returns as measured by these indexes, at least in recent times, so there’s an argument that this isn’t a fair contest. Fundamental indexing should, in theory, offer investors additional return for lower risk.

In a world in which the equity-risk premium over bonds is so thin even small improvements in returns from shares could make a difference to investors. The question is, of course, whether fundamental indexing can really offer that kind of return or whether the models on which it’s based, showing its success in backtesting, will prove to be reliable once the real world gets injected.

We don't know, yet. However, given that a dividend based equity approach would at least have yielded some returns then fundamental indexing strategies which include a focus on sustainable dividend yields deserve some consideration as part (stress "part") of an investor's kitbag.

Bonds and Risk Diversification

Turning his attention to bonds Arnott points out that while using conventional mainstream bonds does reduce our risk they don’t provide the level of diversification we might want. The best performing asset class in the last decade were Emerging Market Bonds, which wouldn't figure anywhere near a mainstream portfolio.

Worse still, bond index funds suffer from the same problem as do market capitalisation weighted stockmarket index funds. In essence, if you track a bond index by market capitalisation then if an issuer sells more bonds your fund must buy more. However, a moment’s thought shows that the more debt an issuer is issuing then the more risky they must be as an investment. The same is true of equity index funds, of course: a stock that doubles in price is not obviously twice as good or half as risky an investment, quite the reverse.

The Return of the Cult of the Equity?

One other argument is that this new research could indicate that shares are cheap. Stocks currently yield more than bonds for the first time since the 1950’s when the Cult of the Equity took off. Back then the move from bonds to equities was triggered by George Ross Goobey, manager of Imperial Tobacco’s pension fund, who recognised that the yield on government bonds wasn’t enough to cover the fund’s liabilities.

Believing shares to be cheap and bonds expensive Ross Goobey switched his investments. In inflation ravaged Britain during the Sixties this proved to be a winning strategy, although it should be noted that Ross Goobey was no equity evangelist. Come the seventies he switched from shares to property, again moving from an expensive asset class to a cheap one.

Arguably the question right now is whether we’re heading for deflation or inflation. If governments keep on printing money then inflation is the only ultimate outcome which is very bad for bonds, although not always particularly good for shares. Of course, these aren't the only asset classes.

New Types of Fundamental Indexing

Arnott's paper is clearly the opening shot in a new fundamental indexing approach, applied not just to bonds but also to other asset types. He appears to be suggesting, and there's more work on this in the pipeline, that to provide investors with safe mechanisms for long term saving it's necessary to develop fundamental indexing vehicles giving access to a whole range of different assets.

If these new instruments give us, the retail investor, the opportunity to invest in new markets and financial instruments while providing real risk diversification then we can at least raise a couple of cheers - even as we survey the tattered remains of our equity portfolios.

Related Posts: Dear Auntie, Why Are My Bonds Bubbling?, Sir Hugh Invents The Share and Gets Lost


  1. Interesting, and having lived through two bear markets in a decade I certainly plan to swap a bit out into bonds next time things get peaky.

    As you say though, that could be some time away. Low valuations and a decade of miserable returns has set the stage for a great run for equities once we get out of the cursed noughties.

    Just to be clear on bonds though, are you saying Arnot has found just *three* 20 year periods where bonds have beaten equities since 1803?! Out of 206 potential 20 year periods?

    Sounds like an amazing argument for equities to me! :)


  2. It's not clear what the performance over the first two periods is actually referring to, but for the latest one it's absolutely clear. Over ever period since 1979 bonds have outperformed equities - so 1979-1999, 1980-2000 and so on.

    The problem is that we have no idea when this is going to end, so as investors we need to deal with it rationally. It's also interesting how supposedly uncorrelated assets all lined up last year - government bonds excepted, a darn good reason for holding some of those. Especially if you're planning to retire anytime soon.

  3. Thanks for the follow up - I thought 3/200 was a bit too good to be true.

    I guess it's the big surprised downwards on inflation rates that did so well for bonds, but I'm still pretty astonished.

    Surely it only makes it look more like we're at the peak of a bull market for bonds... and buying Gilts at 3.5% look even crazier, sans sustained deflation of course.

    On the point of bonds as a diversifier, have you read Swenson? In Unconventional Success, he rejects many assets classes like corporate bonds and hedge funds, but singles out government bonds for exactly the reason you do - diversity in times of fear.