Morgan Stanley Goes Short Treasurys.... Again

Tyler Durden's picture

It has not been Jim Caron's decade. The Morgan Stanley rates strategist, riding on the coattails of the always wrong Morgan Stanley economics team led by David Greenlaw, has been wrong in his annual rates call year after year after year. Which is unfortunate because while unable to see the forest for the trees, Caron does have a better grasp of rates than most other Wall Street penguins. That said, just like everyone else in the status quo, Caron has just come out with another short duration call (i.e. sell bonds), probably the 6th time in a row he has done that in the past 3 years. Perhaps 7th time will be the charm. Amusingly, Caron, terrified to be seen in the same camp as Bill Gross who is short bonds on fears that there will be nobody available to step in an buy the 80% of gross issuance that has been monetized by the Fed to date, make this very loud caveat on his short bond call: "To be sure, our shift toward short from neutral duration has nothing to do with the end of QE2 and related concerns that there will be a lack of  demand to buy US Treasuries once the Fed stops buying them. As we have stated many times in the past, the outlook for the economy will be the main driver of yields, not the end of QE2." No, instead Caron believes that the sell off in bonds will be due to the same bullish economic growth call that he has been predicting over... and over... and over... and over... etc. More interesting is how he suggests the trade is implemented: in MS' view the best way to be bearish on rates is with a DV01 neutral 7s-10s flattener: "we continue to recommend being short 5s on the 2s5s10s fly. In line with the butterfly, and in order to  express a more robust short duration position, we recommend a curve flattener on the UST 7s10s curve: · Sell $133.7mm OTR 7y Notes; · Buy $100mm OTR 10y Notes." Perhaps those who want to be short bonds, but for the right reason, that predicted by Zero Hedge and then Bill Gross, this may be one of the better ways to put the trade on.

More below:

Reducing Duration Exposure from Neutral Toward Underweight

We are reducing duration exposure from neutral toward underweight. The reason is that the market has already well discounted 2H growth to levels much lower than consensus. As we see it, the risk now is for the market to price 2H11 growth higher. We will start out cautiously by expressing this negative view via underweighting the belly of the curve versus the wings rather than positioning outright short. The risk to our view is if the European debt crisis worsens, but our base calls for a temporary reprieve as Greece receives a new aid package. This is a tactical view and we believe the location is good to establish a negative UST bias, given how rich the belly of the curve has gotten and since our models indicate 2.85% is the lower bound of fair-value for UST 10y. Medium term, we believe 10y yields will remain range-bound. Let us explain:

1. The belly of the curve has richened to levels that is consistent with a 2.65% growth outlook over the next six months While we feel it is appropriate for the market to discount the risk of a downgrade to the 3.35% consensus growth expectations in 2H11, we believe that discounting it to 2.65% or lower is too much (growth estimates come from Blue Chip Consensus).

2. Our economics team believes that we will have a rebound in 2H11 led by the auto sector that may contribute as much as 1.5% to growth in 3Q11 (see Exhibit 1, LHS) and possibly as much as 0.5% to 4Q11 growth. Also, consistent with their growth outlook, they see core inflation continuing to rise with headline CPI forecasted to be 3.4% and core at 1.9% by the end of 2011 (see Exhibit 1, RHS). However, this view hinges upon the US economy’s ability to produce at least 150K jobs per month − a key threshold.

3. Over the next several weeks, event risks will come to pass and we expect clarity from the end of QE2, the debt ceiling and Dodd-Frank. We believe that this will reduce uncertainty and the safe-haven bid for bonds. Our macro team believes that the recent soft patch is nothing more than a mid-cycle slowdown and that risky assets may perform starting in 2H11 (see Global Debates Playbook, June 16, 2011).

To be sure, our shift toward short from neutral duration has nothing to do with the end of QE2 and related concerns that there will be a lack of demand to buy US Treasuries once the Fed stops buying them. As we have stated many times in the past, the outlook for the economy will be the main driver of yields, not the end of QE2. Also, we think that an agreement will be made on the debt-ceiling debate between the Democrats and Republicans some time in July before the August 2 social security payment deadline. And as for Dodd-Frank, which is scheduled to start imposing new regulations on the market as early as July 16, we believe that much of it may be postponed until later this year and possibly into 2012 due to the lack of clarity around many of the intended regulations.

Following the money. As the market is priced for slower growth over the past several weeks, we have seen inflows into bond funds rise sharply while risky assets saw outflows (see Exhibit 2). For the month of May, bond funds saw the largest monthly inflows in seven months totaling $20.2 billion while equity funds saw outflows of $2 billion as compared to inflows in April of $5.3 billion (first month of outflows after six consecutive months of inflows). Also, according to surveys we follow, the decline in rates has caused many to reduce their short positions. Thus the combination of money flows and duration surveys we track indicate that the short bond exposure in the market has been reduced which technically puts less pressure on rates to stay low and instead clears a path for them to rise.

Duration risk cuts both ways. When growth expectations for 2011 were being downgraded, longer-duration bonds performed best. This was most notable in TIPS as real yields significantly dropped. The drop in real yields was so dramatic that the year-to-date performance of TIPS even exceeds that of high yield. What has changed? Long-duration exposure presents a greater risk and may become a source for underperformance rather than outperformance going forward, especially if 2H11 growth rebounds as market consensus suggests it might.

Conclusion. Our tactical shift from neutral to short duration has several implications: 1) we expect real yields to begin a steady rise higher, and 2) we expect the belly of the curve to underperform and the 10s30s curve to flatten.

And the best way to express a bearish stance in the rates complex according to Caron:

Since April, the belly of the curve has richened significantly as rates have marched lower (Exhibit 1). We continue to recommend being short the belly vs. the wings, as previously discussed short 5s on the 2s5s10s fly (see “Fade the Recent Outperformance of  the Belly,” US Interest Rate Strategist, June 9, 2011). In line with the butterfly, and in order to express a more robust short duration position, we recommend a curve flattener on the UST 7s10s curve:

· Sell $133.7mm OTR 7y Notes

· Buy $100mm OTR 10y Notes

Both the 7s10s curve flattener and the butterfly allow the investor to play for a reversion in the richness of the belly. Rather than going outright short we suggest initiating these relative value trades, which capture some duration exposure and some relative  value exposure between different points on the curve.

We argue that growth expectations have not been downgraded to the extent that rates in the 5-10y sector have fallen with survey consensus at 3.35% for 2H11. In our view, the rates market is pricing levels of US growth that are inconsistent with surveyed forecasts (see “Reducing Duration Exposure from Neutral to Underweight” in this publication).

This past week, price action in the market has reflected uncertainty as, for example, the 7y point increased 12bp on Tuesday only to decrease by 14bp on Wednesday returning to similar levels. We expect the market to remain range-bound in the near term; however, we see fair value of the 10y note at about 3.10% with a 25bp standard deviation (Exhibit 2). With the 10y dipping to the low 2.90’s, we see an opportunity to fade this extreme.


UST 7s10s Flattener

As we have shifted from neutral duration to tactically short, we seek relative value trades that would perform if yields in the belly increased. We believe a UST 7s10s flattener is one of the best trades that fit this description for the following three reasons:

1) Historically steep curve. We have been in a steep yield curve environment for some time now, but with the recent move lower in yield in the belly, curves such as 5s10s and 7s10s have increased once again and now are nearing the all-time highs reached in November of last year (Exhibit 3). We do not think that the low yields around the 7y point are warranted due to growth expectations higher than they were 8 months ago, and hence we look for this curve to flatten.

2) Beneficial roll and carry characteristics. The 5s10s curve historically has more variance than 7s10s, however both curves have increased approximately the same amount over the last several weeks (Exhibit 4). Additionally, the carry on the 7s10s flattener is -2.0bp per 3m, while it is -4.5bp per 3m on 5s10s. If we divide this carry by the 3m realized volatility of each respective curve, we obtain a carry quotient of -0.24 and -0.36 for the 7s10s and 5s10s flatteners, respectively.

The carry quotient gives us a risk-adjusted level of carry on each curve and shows us that the 7s10s absolute carry of -2.0bp is also better than the 5s10s negative carry on a risk-adjusted basis.

3) Correlation that favors a sell-off in the belly. Our premise is that we are not only fading an extreme curve level, but also gaining exposure to a sell-off in the 7y sector. Recently, that is exactly how this curve has been trading

7s10s has been fairly well correlated with rates since about January, 2010. Starting in October, 2010, this correlation increased, and the beta, or slope of the regression increased as well. Since October, 2010, the 7s10s curve has been flattening/(steepening) approximately 13bp for every 1bp increase/(decrease) in the 7y yield.

The risk to this trade is that the 7y yield decreases relative to the 10y yield. As the trade involves a short position at the 7y point, losses are potentially unlimited.