Now that QE Lite, or whatever one calls it, is here, the most appropriate market strategy reverts back to March of 2009, when life was very simple: "Buy what the Fed is buying." And with the benefit of QE1 in hindsight, namely the Fed's prior purchase of $700 billion in Treasurys in 2009, it is possible to determine precisely which bonds the Fed will focus on, and which are likely to be excluded, thus benefiting the least from the latest bout of monetization. Morgan Stanley has come up with a list of which bonds are likely to be targeted by the Fed in the upcoming 5 Open Market Operations beginning on August 17 and continuing through September 1. Those looking for a quick (and levered) return on investment will be wise to pick up the issues determined as most likely to be monetized, while potentially shorting those that are ineligible for buybacks.
First, focusing on Treasurys which are likely to be excluded, Morgan Stanley's Igor Cashin uses the criteria of avoiding names for which there already exists a heightened demand, such as those that are CTD (Cheapest To Deliver) into September and December future contracts, as well as those trading "special" in repo. Additionally, as the Fed SOMA is limited to owning a maximum of 35% of any given issue, there are quite a few names which are already ineligible for further purchases, as well as many which are approaching the ineligibility threshold. More from Morgan Stanley:
Which issues will be excluded? While the NY Fed has said that it will concentrate its purchases in the 2-10y sector of the nominal Treasury curve, similar to the statement it made at the start of 2009 Treasury QE, it will also buy outside these ranges, as well as throughout the TIPS curve, as we stated above.
The NY Fed will, however, refrain from purchasing securities for which there exists heightened demand. These include the cheapest-to-deliver into the September Treasury future contracts (and probably those of December, as that futures roll is right around the corner), as well as those issues that may be trading special in the repo market (although these issues may be few and far between). To be clear, the Fed has bought on-the-runs before, and these may be targeted again to some degree.
The final criterion for Fed purchases is that they are limited to owning 35% of any single issue, and may also hold back from buying those issues that are very close to this limit. Incorporating all of the above rules, Exhibit 5 displays our compiled list of the bonds that will likely NOT be targeted by the Fed. As a result, these issues may underperform other issues on the curve that are eligible for Fed purchase.
A summary version of the ineligible USTs is presented below: these should most certainly be avoided for the purposes of frontrunning the Fed over the next two weeks, or potentially used in ultra short term and ultra leveraged pair trade combinations.
So which Treasuries is the Fed most likely to purchase? According to Morgan Stanley, those most suitable for "buybacks" (yes, yes, it's not a buyback) are those which are cheapest on average on the Treasury spline for the Zero Vol Asset Swap Curve. In order of upcoming auctions that would mean: the 7.25 of 05/15/2016; the 3.0 of 2/28/2017; the 1.375 of 2/15/2013; the 7.5% of 11/15/2024; and the 1.375 of 1/15/2013. In other words:
Which issues are likely to be targeted? For each operation date, we compile a list of the eligible issues and, after excluding the ineligible bonds per Exhibit 5 (although we leave in the December contract CTDs for now), we rank the bonds that we think are most attractive for the Fed to buy in Exhibit 6, ordered by operation date:
We arrived at our rankings by first calculating the rich / cheap level of each eligible bond off of our Treasury spline, then calculating the average richness / cheapness of that sector by taking the average of those levels, and finally calculating the spread of each issue’s rich / cheap level to the average rich / cheap level of that sector. Eligible bonds are then ordered from highest to lowest. For example, the 7.25% of May16s is the cheapest issue the Fed can purchase in its August 17 OMO, and the 3% of Feb17s is the cheapest issue the Fed can purchase in its August 19 OMO. To help investors visualize the bonds that would be cheapest for the Fed to purchase, we highlight the top five cheapest issues from Exhibit 6 with red circles on a zero-volatility (ZV) asset swap curve in Exhibit 7 below:
Right away, we can see that what looks cheap on our spline also looks cheap on the ZV ASW curve. We can also see that the Fed is likely to target original-maturity 3y notes when it conducts its August 24 and September 1 operations in the 2012-14 year sectors, and is likely to target original-maturity 7y notes when it conducts its August 19 operation in the 2016-20 year sector.
While the Fed may certainly deviate from our rankings, and their ultimate purchases may be driven by additional considerations, the Fed’s QE operations in 2009 did appear to target issues that traded cheap on the curve, showing that they did pay attention to relative value. It is further worth noting that the size of the operations this time will be 30-45% of the average size of the Fed’s UST operations in 2009, which means it may be a bit less effective in correcting the cheapness of certain bonds on the curve.
So in this centrally planned market of ours, why take risk? The Fed is now telegraphing where the free money is, and as such it would be imprudent to avoid grabbing at least some of it. We urge readers to do their homework, but in this particular case (unlike that horrendous steepener trade that Jim Caron just can't get away from, and which we will discuss later) it appears MS is spot on. Yes, the upside won't make one's year, but those who wish to pick a few unlevered bps should have ample opportunity over the next 6 or so months as the Fed proceeds to gobble up a selection of bonds that virtually everyone who has made a lifestyle out of frontrunning the Fed will be gunning after. In other words, this is your chance to front run those who front run the Fed.