Earlier this month there was a series of five articles on market liquidity in the Federal Reserve Bank of New York’s Liberty Economics Street Blog. So much has been written about falling liquidity in the financial markets, with a lot of the blame laid at the feet of regulation. These FRBNY articles discuss some of the other less acknowledged factors, in effect taking the heat off the “unintended consequence of regulation” meme.
Of the five articles, we thought the first and last were the most interesting. In this post we look the FRBNY/LSE first piece, published on August 17th, “Has U.S. Treasury Market Liquidity Deteriorated?”, written by Tobias Adrian, Michael Fleming, Daniel Stackman, and Erik Vogt. This article focuses on how to measure US Treasury liquidity and if those measurements show a sustained reduction in market liquidity.
First, what is liquidity?
“…Liquidity typically refers to the cost of quickly converting an asset into cash (or vice versa) and is measured in a variety of ways…”
The authors examined a handful of methodologies:
Bid-ask spread: “…bid-ask spreads widened markedly during the crisis, but have been relatively narrow and stable since…”
Order book depth: “…measured as the average quantity of securities available for sale or purchase at the best bid and offer prices. Depth rebounded healthily after the crisis, but declined markedly during the 2013 taper tantrum and around the October 15, 2014 flash rally. It is not unusually low at present by recent historical standards…”
Price impact for a given order flow: “…Price impact rose sharply during the crisis, declined markedly after, and then increased some during the taper tantrum and in the week including October 15, 2014. The measure remained somewhat elevated after October 15, but is not now especially high by recent historical standards…”
Trade size: “…After declining during the crisis and then rebounding, trade size also declined during the taper tantrum and around the October 15 event. The decline in trade size compared with the precrisis period may merely reflect the increasing prevalence of high frequency trading in the interdealer market, so may be a less reliable indicator of reduced liquidity than the preceding measures…”
Pricing differentials between on-the-run and off-the-run securities: “…Large pricing differences indicate unexploited profit opportunities, which could in turn reflect constraints on market making capacity and/or poor liquidity…such pricing differences spiked sharply during the crisis, but have been relatively low and stable since…”
Comparing securities with similar credit risk: (the paper looked at) “…the yield spread between bonds of the Resolution Funding Corporation and Treasury securities with similar cash flows…the Refcorp spread also spiked during the crisis, but is currently close to postcrisis lows, albeit somewhat above precrisis levels…”
(Note: when was the last time anyone even thought about RFC paper? Maybe it was the cleanest comparison available, but this is really out of left field.)
Most of these indicators are disputing that there is less liquidity now.
“…The evidence suggests that market participants’ liquidity concerns are not emanating from average levels of liquidity in the benchmark Treasury notes…”
So why is it that market participants have been raising the flag? The authors wonder if the concern is focused on off-the-run US Treasury securities or other fixed-income securities like corporate bonds? That is certainly part of it.
Alternatively, they postulate that
“…the concerns are not so much about average liquidity levels, as we examined, but about liquidity risk. Indeed the events of October 15 and similar episodes of sharp, seemingly unexplained price changes in the dollar-euro and German Bund markets have heightened worry about tail events in which liquidity suddenly evaporates…”
and finally, the post suggests maybe the fear is about future liquidity concerns and imbalances between supply and demand.
“…On the demand side, the share of Treasuries owned by mutual funds, which may demand daily liquidity, has increased. On the supply side, the primary dealers have pared their financing activities sharply since the crisis and shown no growth in their gross positions despite the sharp increase in Treasury debt outstanding. Market commentators point to these factors and the current environment of low volatility and worry about what will happen when monetary policy is normalized and volatility rises…”
(Note: We wonder about the averaging that the charts use and if they aren’t crafted to the message? Most are 21 day moving averages. This makes the charts much more readable, but will also smooth out the data just when you want to see exactly how episodic liquidity can be. Average liquidity is not the problem.)
The liquidity indicators examined certainly went crazy during the financial crisis and the flash crashes/taper tantrum. But the message is that everything is ok now. This post, and to a large extent the others in the series, present a story of “move along, nothing much really happening here”. Anecdotal evidence we have heard/read is in the other direction.
Speeches on the topic, and in particular Governor Powell’s August 3rd speech entitled “Structure and Liquidity in Treasury Markets” (we wrote about it in an August 4th SFM post “Fed Governor Jerome H. Powell spoke on the factors behind disappearing US Treasury liquidity. Its not just regulation.”) echoed this sentiment in the posts. Powell cited improved technology & high frequency trading (see Michael Lewis’ “Flash Boys” for a primer) and those topics are revisited in the Liberty Street blog posts. Regulation isn’t left out of the mix, but the Fed seems to feel it is not the driver that many financial writers would have you believe. Is this spin? Or are the Fed economists and researchers telling us the facts simply don’t support the popular notion that illiquidity has permeated the US Treasury market? Color us skeptical.