I like challenging widely held assumptions. That’s how you can outperform anyway, or at least, differentiate. It’s also much nicer to try and think differently.
Today I want to challenge the widespread faith in the balanced, 60/40, “risk-parity” portfolio. That is to say, the one with equities and bonds balanced to offset risk. Most people believe in diversification – using bonds to offset some of the risk from equities.
I want to try and convince you that the 60/40 bond correlation is based on ideas which will not last forever, and that simply using bonds is insufficient diversification against true risk.
(What is “risk” in investment? Bloody hell – I’ll answer this one in full another time, but my favourite so far is that there are two primary risks – the risk of losing money and the risk of not making money. In financial markets, volatility, or the “Vix”, is often thoughtlessly used as a simple stand-in for “risk”, which is part of the problem. In this case, I meant the risk that things do not go as you expect, leading to either of the two primary risks – loss, or absence of gain.)
The assumption of such portfolios is that the equities and bonds are anti-correlated, which means that when stocks go down, bonds go up.
Stocks go up more than bonds in the long run, because of their two-factor compounding (reinvestment of dividends to increase your holding, and investment of retained earnings in growing the business).
The 60/40 portfolio needs semi-regular rebalancing, to keep that balance in check.
Here’s what happens to a 60/40 portfolio over time if it is not rebalanced (equities sold and bonds bought to restore the balance when stocks outperform).
Source: The Irrelevant Investor
What you see there is that equities have outperformed so drastically over the long term that they erode the bond holdings down to a tiny proportion, simply by growing so much.
But if stocks have always outperformed, what’s the rationale in investing in bonds so blindly?
Well the idea is that bonds offer a diversification and thus a lowering of risk (I think in this case it means risk of loss). Bonds offer a fixed, predictable return, and tend to rise when equities fall, thus offsetting some of the worst effects of a stockmarket crash.
The underlying assumption is that equities and bonds do not correlate. And it is one I’d like to challenge today.
I wouldn’t go as far as saying it’s wrong, but it’s not 100% right and I think many people have placed too much faith in it.
Despite what the investment textbooks say, it is not a straightforward anti-correlation. It has been in recent decades, but that is not the same thing. It’s a trend, not a law.
In fact, since the 1870s, equity-bond five-year correlation has been negative in 635 months compared to 1,063 months when it was positive. As in, out of roughly 1,700 months, over 1,000 of them have seen equities and bonds move in the same direction.
A Bank of England analyst has collated data from various sources and shown that in the UK, for 250 years prior to the current anti-correlated relationship, stocks and bonds generally moved in tandem to various degrees:
Source: Man Group
If the blue line is above zero, that means stocks and bonds are somewhat correlated – ie, they move in the same direction. You can see there that the trend towards and into anti-correlation started at the beginning of the 1980s.
It is precisely 0% coincidence that this also marked the beginning of a number of other phenomena, which in my mind are all connected: falling interest rates, falling inflation, incredible stockmarket returns, a compounding bubble-bust cycle, increasing debt levels, and increasing market dependence on central banks, which have acted as a fire-fighter and arsonist in equal measure (to steal a metaphor from this excellent summary of central bank policy over the last 20 years).
Why then is the stock-bond anti-correlation idea so prevalent?
I certainly learned it that way round first, and have only slowly come across challenges to that assumption later.
The reason, according to Chris Cole of Artemis Capital, is that funds and investment firms don’t want to use anything other than perfect data, and that only really exists in a computerised form since the 1980s. That means all the models they run only go back a few decades.
He points out that the average wealth adviser (in the US I’m guessing) is 52, and was therefore in primary school the last time inflation and rates were rising. Even the slightly older ones will have learned their trade in deflationary era of falling rates, incredible returns, falling volatility, and inverse stock-bond correlation, which explains the prevalence of the assumption.
That’s the first assumption I want to challenge regarding the very common belief that the 60/40 portfolio is “less risky”.
To take a closer look at the bond part of the portfolio specifically, there are new challenges that they face.
Remember, they clearly offer an inferior long-term return to investors, and seem to only have a place in the 60/40 balanced funds because in theory, they reduce risk, often lazily measured by volatility of price.
But we do not live in theoretically normal times. In fact, we live during a period of unprecedented monetary policy. Never before have interest rates been so low, or for so long.
And this causes a problem for bonds. Bond yields – a reflection of the coupon relative to the price (yield = coupon/price) – move depending on what happens with a national interest rate.
That’s because the value of a bond’s coupon (£2 per year for a £100 bond) is measured relative to a national interest rate, as well as how long before you get your “principal” back.
£2 per year per £100 looks useless if national interest rates are at 3% and your bank can do better, for example.
But with interest rates so low, a 1% change in interest rates makes a much bigger difference than when national rates were higher.
If a bond yield moves from 1% to 2% then its price is cut in half. A doubling of yield means that £2 is out of a £100 bond price rather than a £200.
If yield moved from 5% to 6%, you can see how the relative change would be smaller and the change in the price of your bonds would be smaller too.
So, with national rates so low, any move in that interest rate has a much larger impact on bond prices than before.
This means increased bond volatility, which is often used as a proxy for risk.
So suddenly, the “low-risk” portion of your investments is now much more volatile than normal, despite yielding much less in terms of income.
And I hate to break it to you, but there’s another problem too.
As rates have slowly fallen towards zero, the available return from bank accounts and bonds has fallen, leaving many funds “searching for yield”. Imagine a pension fund that planned its future payment on the basis it could achieve a safe, 7% annual yield.
But with rates next to zero, that’s much harder to achieve, as the 40% of your risk-parity portfolio isn’t delivering its fair share.
So what do they do? They use leverage to get that extra yield. You use debt to supercharge your returns, just to eke out the extra 1% or 2%.
But leverage is a pact with the devil – it grants you additional powers for a while, but when they call in their payment, you’d much rather be dealing with anyone else.
It may offer exaggerated returns on the way up. But it does the same on the way down, and as we all know from recent memory, markets creep on the way up, and freefall on the way down.
So now the “safe” part of your 60/40 portfolio has been made less safe.
It’s more volatile, and it’s using more leverage.
And the assumption of anti-correlation it was built on was a lie.
And all of it is based on three underlying factors: falling rates, low inflation, and falling volatility.
A regime has existed since the 1970s, when the fed funds rate reached its 200-year peak of 19% under the stewardship of Paul Volcker, who did so to rein in rampant inflation.
Since then, rates have been chiselled and occasionally hammered lower, to lie now on the floor.
Source: Federal Reserve of St. Louis
I’ll leave it there today, having focused on the bond aspect of the 60/40 risk-parity portfolio, but on Wednesday I’m going to try and take it a step further.
Firstly, I want to look at what happened in the March crash, to see some of this in action. I also want to look at some more current, macro reasons for why this anti-correlation could be coming to an end.
But most importantly, I want to investigate what you could consider for true diversification, which allows for a change in the very rules of economics as we currently believe them to be. But they are not rules, merely trends, and they could be nearing their end.
Have a great week everyone,
Editor, UK Uncensored