More on GS and the relevance of safe assets | FT Alphaville

More on GS and the relevance of safe assets

Izzy already did a great job giving the house sentiment about the dismissive Goldman Sachs note on safe assets, but we wanted to add a few quick points.

This isn’t just gratuitous piling on or a therapeutic exercise or a reflexive defence of our earlier posts. To the extent that others hold what we believe are the note’s misconceptions, it’s best to discuss them out in the open, loudly if need be.

Some of this will be repetitive to what we have previously written, but a few points in no particular order — and again consider this a follow-up to Izzy’s earlier post, which already tackled some of the other major issues:

— Goldman seems to think it is minimising the relevance of the matter by arguing that supply hasn’t fallen, but rather simply hasn’t kept pace with demand.

But this is precisely why it is so important. When this shortfall happens in times of financial stress, it means that central banks and fiscal authorities are not doing their jobs. (When have we ever said that the decline in absolute supply is what matters?)

To repeat ourselves, it means that households are hesitant to spend their incomes; that companies are less likely to invest retained earnings; that banks will be hesitant to extend credit (both absolutely and relative to demand for it) given the lack of available collateral — clogging traditional monetary transmission channels; more emphasis will be placed on counterparty reliability, leading to further concentration in the financial sector; repo markets will remain strained; and investors will be loathe to diversify across risky assets.

We didn’t just make this stuff up. Some issues specific to shadow banking and credit intermediation are recent developments of the past few decades, but the basic principles have been the same since Bagehot wrote about them in 1873.

— Goldman argues that it is trying to bring nuance to the conversation, then lumps all safe assets together regardless of locality to make its point.

For instance, it’s hard to tell from the note, but Goldman seems to have included Japanese government bonds in its total calculation of global safe assets. The amount of JGBs outstanding has increased, according to a back-of-napkin calculation using current market exchange rates, by more than $2 trillion worth since 2006. (See here, page 7.)

And hey, JGBs are safe assets. But they are also mostly bought by domestic investors and, more relevantly, what good does the increased issuance of yen-denominated debt do for credit intermediation issues in Europe and the US? (And these economists accuse others of being simplistic?)

This might seem like nit-picking, but we think it reveals a broader misunderstanding on Goldman’s part.

And if it doesn’t, then what’s their point? Why use the fact that the global supply of safe assets has increased to diminish the shortage story when those of us who have been writing about it have never denied that locality (and more precisely, the specific currencies of specific funding markets) matters?

We’d also mention that when safe assets have money-like properties, as those denominated in US dollars do, a simple return to the absolute level of safe assets rather than their growth trend is problematic in times of crisis when demand has shot up.

— Goldman writes: “We think the largest effect comes from the ongoing deleveraging process and the broad weakness in this post-bust recovery. The increase in government supply is itself a part of this process and the demand for safe securities is the other side of the same coin, not a separate force. If these pressures ease, the demand for safe assets is likely to fall, but so is the supply.”

This is wrong.

When the supply of safe assets is insufficient to meet demand, we’re probably in a recession. When the supply of safe assets outstrips demand, we have a bubble.

In the current situation, we have a big shortfall. So if demand goes down because deleveraging slows and the economic recovery accelerates, it’s not immediately obvious that supply will also go down given the gap that now exists between the two.

It is much more likely that when deleveraging slows and the economy recovers more quickly, some of the production of safe assets done by the government will be replaced by safe assets created by the private sector. These assets are still needed in the current financial system, and the point is that the equilibrium described above is what matters.

The danger is exactly that given the role of safe assets in current financial markets and their money-like properties, governments and central banks will tighten policy too quickly before the private sector is growing fast enough to replace the public-sector production of safe assets.

Someday we might live in a world where safe assets don’t play a meaningful role in the financial system, or at least play a less meaningful role. Probably such a world would be, erm, safer and less prone to crises. But it’s unclear that anybody has worked out how to get there, and at any rate we’re not there now.

Or if you want a quicker response, consider that safe asset supply was rising, not falling, in the years leading up to the crisis. Have a look, for instance, at the charts in this post.

— Goldman pays lip service to the issues of asset encumbrance, collateral rehypothecation and the money-like properties of safe assets, but doesn’t find these to be a very important part of the story.

We’ve written a ton about this and simply disagree — repo markets and the move towards central counterparty clearing mean that this remains a very big deal indeed. We’d refer you instead to our earlier two posts, to the IMF and to Singh/Stella.


The Goldman economists aren’t idiots and certainly not everything in their note is wrong. Their overall argument is summarised near the end: “…it is not clear that the shortage of safe assets is truly represents a separate policy problem to be solved.”

Maybe not, but it’s a wide-encompassing framework through which to understand everything and makes an awful lot of sense — a framework, by the way, that was poorly understood by economists everywhere before the crisis, and remains under-discussed even now. We find it strange that Goldman would be so flippant about it.

So Izzy is right: Goldman showed up late to this party, but that’s no excuse — the IMF did too, but their effort was much stronger. It’s also why we prefer the work of Sweeney and Wilmot, who have been investigating this since 2009.

So while it’s good that Goldman is looking into the matter, we hope next time they’ll look a little harder.

Related link:
Goldman Sachs on fewer ‘safe assets’ – FT Alphaville