While we often pick on Morgan Stanley's Jim Caron (the same guy who year after year after year keeps predicting the yield on the 10 year will soar, and not just soar, but soar for all the wrong reasons, such as bull steepening and what not), has just diametrically changed his tune, by bringing us, drumroll please, Operation Torque. To wit: "Policy makers in both the US and Europe get back to work in September, and this month will be rife with deliberations on stimulus and market support policies. In our view, a duration extension to the Fed's SOMA portfolio is an optimal policy tool to engender easing. This can initially be done through extending the duration of reinvestments from MBS and agency holdings but may ultimately culminate in selling shorter-duration USTs in its SOMA portfolio in exchange for buying longer duration assets (‘Operation Torque’, as we at Morgan Stanley have dubbed it)." Why 2 Years? Because as per the August 9 FOMC statement, we know that there will no rate hike for the next 2 Years, and hence no duration risk. Which means that the Fed can sell an infinite amount of paper into a mid-2013 horizon without worrying about demand destruction. And by doing so it will, as we have been predicting since May, expand the duration of its portfolio, in the process pushing investors into risky assets for the third time in as many years. But there is a twist...
...Or a torque as the case may be...
As we have noted on several occasions, going aggressively after the 10 Year would likely mean an even more pronounced flattening of the 2s10s than we currently have: an outcome which more than anything will impair US banks. As yesterday's MBA mortgage refinancing data showed, even at record low mortgage rates virtually nobody wants to refi. Perhaps a forced refinancing will help courtesy of Obama, but we think not. The point being that net interest margin, or the carry trade as it is better known, will all but disappear, and perversely QE3, call it Twist or Torque, will end up generating even more pain for the critical financial sector, without which there is no chance of a broad market rally.
Furthermore, we are amused that MS is implicitly agreeing with us: Caron says - "We believe that if the Fed implements Operation Torque, it will purchase a significant amount in the 30y sector of the curve as well as at the 8-10y point." And it continues: "If the Fed does purchase on the long end, the belly could easily underperform with yields currently near all-time lows. We recommend taking advantage of this opportunity by positioning for the 7s30s to flatten, or trading 7s30s on asset swap as described below." In other words, a flattener with a 10 year hump, so needed for even some incremental 2s10s steepening... Just as we have been predicting.
However, that's not all. We are confident that when all is said and done, the Fed will realize it needs to drop the 10 Y yield modestly in order to afford some profitability for the banks, resulting in an emphasis on the Ultra longs (17-30 Year sector).
There is however a glitch: there is nowhere near enough supply in the 17-30 Year space to meet this need for 10s30s flattening even as the 10 itself floats higher. And while 10 is too short, the 30 is too long, does this mean that the Treasury will soon have to consider converting the 20 Year point on the curve from a simple Constant Maturity Series into an actual cash bond to satisfy the suddenly very picky Goldilocks environment ? Unless Geithner wants to take a chance with flooding the market with 30 year paper, for which the only buyer will be the Fed, it may soon have to.
Anyway, back to Morgan Stanley's Operation Torque. Here is how Caron presents the three core scenarios that the Fed will undertake:
The goal of Operation Torque is to remove duration supply from the market, and not simply to push yields lower. With less supply in the market, risk premiums for spread products should decrease, driving easier financial conditions.
During QE2, the Fed removed $490bn in 10y equivalents form the market. If we use that as a baseline, we may evaluate Operation Torque under several different scenarios. As the Fed would not expand its balance sheet under an Operation Torque, there is a limited supply of short-duration debt with which to extend, and the market impact is highly dependent on where it is placed. The three scenarios below help us understand targeting purchases on different parts of the curve could impact the market.
Scenario 1: Fed sells all debt with less than two years to maturity and allocates proceeds to the 8-10y sector.
Scenario 2: Fed sells all debt with less than two years to maturity. Proceeds allocated to the 8-10y sector and 30y sector, targeting a total of $400bn equivalents removed from the market.
Scenario 3: Fed sells all debt with less than two years to maturity. Proceeds allocated in same ratio as QE2 purchases.
All scenarios assume a time frame through January 2012 and include estimated auctions. We only include USTs in this particular analysis; however, we estimate that mortgage pre-pays could add up to an additional $90bn should the Fed use those funds to extend duration in the SOMA as well.
And once again, we don't like to gloat, but we did "tell you so" - the Fed will target not the 10 year but the 30 year:
By purchasing in the 30y sector in addition to the 8-10y sector (scenario 2), the Fed could achieve a $436bn purchase of 10y equivalents, having sold the same debt out of the SOMA. This is very close to the same level of duration removed from the market during QE2.
Finally, we may compare to an operation that has the same allocations as was done in QE2 (scenario 3). This style of operation has the least impact in duration terms at $186bn – less than 45% of the impact of Scenario 2 (Exhibit 5). Additional detail of each scenario can be found in the appendix to this report.
We believe that if the Fed implements Operation Torque, it will purchase a significant amount in the 30y sector of the curve as well as at the 8-10y point. In order to position for this Fed action, we recommend the following:
- Overweight 9-10y & 30y UST sectors
- UST 7s30s yield curve flattener
- 7s30s asset swap curve flattener ? UST 7y to underperform and 30y to outperform versus Libor
We look for trades that benefit from duration buying at the 9-10y and 30y points on the curve. In particular, we recommend duration longs to re-allocate into such as 2020 and 2021 original maturity 10s and 2041 bonds as these have a lot of par available to the Fed and are likely candidates for purchase.
Bottom line: Goldman, JP Morgan, Nomura and now Morgan Stanley all assume QE3 is a fait accompli, the only question is what shape it will take.
And for all intents and purposes, what the Fed will achieve, is to get investors to rush out of anything with a sub 10 Year duration, and into the longest point on the curve. And just like last time, the biggest beneficiaries will be not bonds, nor stocks, but commodities, where the marginal purchasing power is far greater, and the result will be yet another round of geopolitical shocks, this time, as we have said so many times before, far closer to the core: both in Europe and the US. As for the effectiveness of such a move on the economy and stocks, we urge readers to look at the following chart.
Below is the complete Morgan Stanley moral hazard playbook.