Safe assets are bunds, treasuries, gilts, JGBs and other government bonds issued by large developed economies. In one sense, they actually are ‘safe’ assets. They are typically negatively correlated with risk assets. More specifically, if the risk of recession is perceived to be rising, ‘safe’ assets rally. That’s a good reason for a zero term premium – why should you get paid to insure against recession?
Of course, for serious, long-term investors – Buffett, Soros and anyone else with a long memory, profound thoughts and a proper time horizon – these are not really safe assets. First of all, they are priced to deliver negative real returns, and the skew on the return distribution is all to the downside. The best an investor can hope for is to lose a little … when the regime changes, you rarely get a heads-up.
But that’s an aside, this isn’t a blog about risk. It’s about shortages. A recent twitter exchange with Brad DeLong reminded me of Fischer Black’s masterpiece, Noise. The answer to any financial market anomaly can be discovered therein.
One of the many brilliant insights is into the subject of liquidity. Fischer Black observed that noise traders do have a social purpose, they create liquidity, and liquidity reduces transaction costs. That’s a reason why high frequency noise machines are not all that bad: they make index-trackers cheaper.
Anyway, that’s a distraction. I have had a problem with the phrase ‘shortage of safety assets’ as an interpretation of global bond pricing, for some time. So I tried to imagine if Fischer Black would buy the idea. I don’t think he would.
In a highly liquid market the concept of ‘scarcity’ is meaningless. The term ‘scarcity’ is a freudian slip, it has associative wisdom, because in evolution scarcity motivates. Food was (sometimes is) scarce. Ask children if they want a slice of cake; if they don’t, tell them it’s the last one. If they still don’t, tell them you’ll give it to their sibling.
And so it is with bonds. Is ‘scarcity’ an ex post rationalisation to explain an evolved propensity to panic, which is dysfunctional in investing?
Bunds, treasuries, gilts et al, remain among the most liquid assets in the world. You can buy tens of billions of dollars worth, at will. When there is a shortage of something – beer for example – you can’t buy it, even if you want to, at any price. Yes, beer can be illiquid, too.
So what is going on in bond markets? Something far simpler: bond prices in highly liquid markets are determined by interest rate expectations and ‘risk’ preferences and behavioural biases. One of the most dangerous behavioural biases is the fear of a shortage …
Far from being a proof of anything, this blog will strike many as a glib set of pronouncements. Is there a test of this hypothesis? To be clear, I am not discussing fractional changes in bid offer spreads and anomalies in on- and off-the-run treasuries caused by the increased regulatory cost of market making. I am talking about the collapse in bond yields and the elimination of term premia. Well, let’s see what happens to bond prices if the equity risk premium falls and interest rates rise. I predict that bonds will fall significantly, what does the ‘shortage’ thesis predict?
Finally, I should clarify that these thoughts have no bearing on my perspective on fiscal policy. Governments should be issuing ‘safe’ bonds hand-over-fist to fund investment in R&D, infrastructure and human capital. I am firm believer in the ‘Cristina Romer doctrine’ in fiscal policy:
“And when will it be time to think about long-term budget balance? As I believe my colleague Christina Romer used to tell you every single week: the bond market and the inflation rate will tell you when it is time to turn to dealing with long-term budget balance. They are certainly not telling you to do so now.” – Brad DeLong
Excellent descriptions of the changing nature of financial intermediation are provided by Izabella Kaminsky here, here and here again. This from Frances Coppola is also, as always, excellent. My interpretation is that Izabella and Frances are not really making observations on global bond yields, but rather about the changing nature of money within the financial system, and the non-bank financial sector’s ability to expand its balance sheet. The connection between their observations and aggregate demand, is, to me at least, somewhat opaque. If I am wrong about this, I have no doubt they will explain why.
Thanks also to James Mackintosh for a series of exchanges on liquidity and shortages.
A very insightful Twitter exchange followed this post – focusing on the whether it is possible to have very high levels of liquidity in a market with shortages, and how one would test the shortage hypothesis. I remain firmly of the view that high levels of liquidity are inconsistent with the idea of shortages. Also, ‘shortage’ cannot be defined simply as ‘too high a price’. Anecdotally, there are two major bond markets where there is periodic commentary bemoaning diminished levels of liquidity – the gilt market and JGBs. This evidence on the JGB market from the BoJ is interesting. There is some evidence here to support a reduction in liquidity. To my knowledge there is no similar evidence at all in Treasuries.