One of the notable observations in recent Treasury auctions has been the increasing participation by commercial banks in taking down 10 and 30 Year Treasury auctions - traditionally two parts on the curve banks have historically avoided like the plague. We present some observations on why this may be happening as well as some troubling conclusions, both of which indicate trouble, namely that liquidity is and has been the name of the game for the past 13 months, and that commercial banks, or presumably some of the smarter money around, are seeing economic distress ahead.
Here are the preliminary observations from Morgan Stanley, the one outlier bank which sees rates jumping to 5.5% by the end of the year, and 4.5% by the end of June (by way of context):
Historically, commercial banks tended to purchase Treasuries in the 2-5y sector, with little interest beyond that. This has been because banks are in the business of taking on risk, not parking their cash in a risk-free asset. And in a post-recession world (e.g.,
loan demand is down, net issuance of MBS is negligible), banks have tended to purchase in this sector as a way to earn net interest margin (NIM) while waiting for a recovery, with little need to take on additional interest rate risk beyond that.
However, recent auction data indicates that commercial banks – defined by the Treasury to include “…banks, savings and loan associations, credit unions, and commercial bank investment accounts” and wholly separate from broker / dealers – have, for whatever reason, started to purchase Treasuries in the back end. Beginning with the February auctions, banks bought $2.7 billion 30s, and in March auctions bought $2.6 billion 10s and $3.1 billion 30s (April allocation data is not yet available). Prior to this, banks had never bought more than $1.6 billion and $0.4 billion in the 10y and 30y auctions, respectively. We do want to stress that buying 10s and 30s is very atypical for banks and is something we do not expect to catch on going forward. Exhibit 4 demonstrates this recent trend in the auction allocations for 10s (middle) and 30s (right):
Why is this occurring? We find MS' explanations both compelling and troubling.
A number of explanations have been discussed for this latest trend, and while each topic could lend itself to a lengthy discussion, below we only briefly highlight the issues. Possible explanations, all of which indicate that the market is increasingly detached from reality - a recurring theme on Zero Hedge, include:
- Banks are increasingly bearish on the economy and, similar to Japan, are willing to buy longer-dated Treasuries. We point out that this conflicts with the continued improvement in bank earnings (e.g., JPM reported 1Q10 EPS of $0.74, beating the MS estimate of $0.48 and consensus of $0.65). According to our banking analyst Betsy Graseck, this is leading banks to position for a “less bad” consumer and rising rates, not falling rates.
- Lack of alternative securities supply leads to a purchase of Treasuries. The option-adjusted duration (OAD) of a current-coupon MBS is 6.2 years, nearly an exact match to the 6.17-year duration of the 7y Treasury. Therefore, we would have expected banks to possibly purchase 7s as a substitute for the lack of MBS supply, while the bank bid continues to be nonexistent in the 7y auctions (Exhibit 4 – Left).
- Deposits continue to flow in and need to be invested. The rise in deposits has to be invested – a fair point. We argue that banks are much more likely to invest in the 2-5y sector where they can still earn a high NIM. Banks might be able to forecast the level of fed funds out a few years (and hence project their NIM), but out to 30 years? Not at all likely, in our view.
- Adoption of FAS 166/167 means banks are taking on low-duration card loans and could be trying to extend the duration of their securities portfolio to offset this. The need to re-extend the average duration of their assets, banks could use a barbell strategy by buying 30s. But given that their actual exposure did not change – they were always exposed to low duration loans, just previously off balance sheet – why make the cosmetic change now? In any case, this would seem to point to a one-off event.
- Reallocation to long duration / away from short duration is an implicit flattener hedge. A possible explanation goes back to what we know is true today: the yield curve is steep and NIM is at its all-time high. So while banks could park their cash in the traditional 2-5y sector, a few may be underweighting this sector and buying 30s instead. This functions as an implicit flattener – underweighting front end / overweighting back end – and helps hedge their reinvestment risk.
- Back-end swap spreads are inverted, making longer-dated Treasuriesattractive. Continued inversion of back-end swap spread – although we do not think they revert anytime – also makes back-end Treasuries more attractive to receiving swaps. This adds to the point directly above, making it more attractive to buy back-end Treasuries than receive 30y fixed.
Net, we stress that we need more data before we can conclude if this is indeed a start of a new trend or that it’s even important to the level of back-end rates. After all, 30y yields have continued to trend higher even despite these auction allocations (Exhibit 5). We also stress that the total amount of Treasury purchases by banks continues to pale in comparison to supply, and despite this recent oddity, 30y yields continue to trend upwards. Therefore, we do not anticipate changing our core call for higher rates and steeper curves based on this recent auction-allocation data.
Flatteners? Excess Liquidity? Gloomy economic outlook? The presented potential culprits are all glaring red signs that either the fundamentals are in overdrive and the market has overshot the bounceback, or that the fundamentals are completely irrelevant, as liquidity prevails, and nothing really matters, except for the actions of the Fed: a primary theme on the pages of this blog. It seems that by presenting the liquidity theme as an increasingly prevalent one to pervert economic reality, major banks such as Morgan Stanley are slowly sending a message to the Fed - yes, Ben, it is blatantly obvious that there is a bubble. Yet nothing matters until Goldman confirms this. And we know that Hatzius' golden boys don't see a rate hike until 2012. In which case the risk/return profile of an equity investment, where "regulatory" risk now prevails in the shape of what side of the bed Bernanke wakes on, is congruent to that of "investing" one's money in the craps tables at the near bankrupt MGM Mirage (which due to its massive short interest was upgraded from Buy to Conviction Buy earlier this week by Goldman). Judging by how much fun trading has become, we would recommend that everyone go to Vegas stat. At least the drinks will be comped...