What does it mean to be “diversified” when everything is expensive?
That’s a good question. Does “diversification” help when everything is in a bubble? Maybe at the margin, right? I mean, “diversification” entails holding assets the returns on which are negatively correlated but when all of those assets are in bubble territory, the implication is that at best, returns going forward are going to constrained and at worst, returns will be negative. So while being diversified across a bunch of assets that are all expensive might help mitigate exposure to one of those assets plunging, the concept of “diversification” becomes more ambiguous in an environment where everything is overvalued.
As Goldman writes, in a new piece out Tuesday evening, “valuations across assets are expensive vs. history, which reduces the potential for returns and diversification [as] valuations become a speed limit for returns.”
Hopefully, you can already see where this is going. A combination of factors have conspired to push valuations across assets higher, even as we are in the midst of the longest period of negative stock/bond return correlations this century.
Equity and bond returns have become negatively correlated since the late 90s – the last 20 years has been the longest period of consistently negative equity/bond return correlations in the last 100 years (Exhibit 10). 1987 was the first year when equity/bond return correlations turned negative; equities had a sharp drawdown (‘Black Monday’) and the Fed responded to financial market stress by cutting rates and providing liquidity – it was the birth of the ‘central bank put’. It has also become more credible due to the more anchored inflation expectations as a result of the ‘Great Moderation’ and as central banks have buffered growth shocks more proactively since the 90s in their quest for ‘inflation targeting’
Between that and the occurrence of low vol. regimes, risk-adjusted returns for a 60/40 portfolio since 1985 have been consistently better than the S&P.
Note the CPI line in the chart above. Persistently low inflation has been the catalyst for a rally across assets as central banks’ efforts to combat the disinflationary impulse drive the search for yield, fuel carry trades, and critically, support equities. As Deutsche Bank wrote earlier this year, “bonds are expensive to equities and by staying expensive can allow equities to become more expensive – bonds thus serve as the crutch to the equity market.” Here’s Goldman again:
And elevated valuations increase the risk of drawdowns for the simple reason that there is less buffer to absorb shocks. The average valuation percentile across equity, bonds and credit in the US is 90%, an all-time high. While equities and credit were more expensive in the Tech Bubble, bonds were comparably attractive at the time. The current valuation percentile is most comparably to the late 20s, which ended in the ‘Great Depression’, and the 50s.
Yes, “the current valuation percentile is most comparably to the late 20s, which ended in the ‘Great Depression’.”
And that’s not the only thing about the current state of affairs that’s “comparable to the late 20s.”
As it turns out, this is now the longest stretch without a 10% drawdown for a simple 60/40 portfolio (which of course depends heavily on the notion of a negative stock/bond return correlation for its legitimacy and for its claim to being the cornerstone of a prudent long-term investment strategy) since the Roaring 20s. Here’s Goldman again:
We are nearing the longest bull market for balanced equity/bond portfolios in over a century – a simple 60/40 portfolio (60% S&P 500, 40% US 10-year bonds) has not had a drawdown of more than 10% since the GFC trough (8.7 years) and has delivered a 143% return (11% p.a.) since then. The longest run has been during the Roaring 20s, ending with the Great Depression. The second longest run was the post-war ‘Golden age’ in the 50s – the 90s Boom has been in third place but is now fourth, after the current run.
Obviously, this most recent run is attributable in large part to stocks, but don’t forget that bonds are in a three-and-a-half-decade bull market, which means that both components are contributing. That, as the return correlation has remained consistently negative thus providing both diversification and double-barreled performance.
The question here is the same as it always is when anyone has this discussion these days. Namely: what happens in a tantrum scenario where a selloff in bonds is the trigger event that leads to a spike in cross-asset volatility and spills over into stocks, thus flipping the return correlation sustainably positive? And wouldn’t this be all the worse given valuations and increased duration risk? One more time, from Goldman:
At current low yield levels and with rising inflation, bonds might again be less effective hedges for equities. Compared to the 90s, they offer less carry and repricing potential to buffer ‘growth shocks’ following the multi-year bond bull market. At the same time, they have become a source of risk in case of ‘inflation/ rate shocks’. And gradually rising inflation could move the central bank put ‘more out of the money’, requiring a larger ‘growth shock’ for central banks to ease policy due to limited options. As a result, equity/bond correlations might become less negative and risk-adjusted 60/40 returns might be lower.
From where we’re sitting, the more likely (and far more dangerous) scenario is an inflation shock and a knock-on policy shock. That, as opposed to a growth shock which would presumably dent stocks, but not bonds. While Goldman notes that “a bear market for bonds alone does not have to result in a large, lasting drawdown in equities,” it seems entirely reasonable to suggest that the prospect of a less predictable policy put (which could very well be the result of an inflation shock), would be the proverbial death knell for the equity rally. At that point, the “nowhere to hide” scenario comes into play.