Rebalancing

Macro Man is back, belatedly, after a glorious few days of eating, drinking, football, and even a little cycling.  He returns to what can be described as the business end of the year, with a lot of important event risks to navigate before books are taken down for the holidays.

Before considering payrolls, the Italian referendum, ECB, and Fed, however, we must also be cognizant of how far markets have come recently.  US fixed income markets are on course for their worst month since January of 2009, and the fourth worst since at least 1988.  While there has been a lot of shuffling of winners and losers under the surface, on an index basis US equities have done quite well as hopes of Trumpflation have gripped markets.

Put it all together, and you have the sixteenth biggest monthly SPX outperformance of Treasuries (on a total return basis) since 1988.

Now, this might naturally lend itself to concerns that we'll see some month-end rebalancing from the pension guys who suddenly find themselves overweight equities and underweight bonds.   Today's price action thus far fits that narrative to a T.   However, does this sort of behaviour really drive bond price action?   Macro Man decided to investigate.

As there are three trading days left in the month (including today), Macro Man parsed US 10 year yields by their movement over the last three trading days of every month since 1988.  He then sorted this data by the amount of aggregate equity outperformance over the course of the month, and focused on those months where stocks either over- or under-performed bonds by at least 5%.   Ifg there is something to the rebalancing gig, we should expect to see Treasury yields dip when stocks have outperformed and rise when they have underperformed.

As it turns out, the size of the sample window and the size of the long bond bull market mean than on average, bond yields have dipped no matter how you slice them.   However, the evidence seems to suggest that the last 3 trading days of the month are characterized more by momentum than mean reversion.  As the table below illustrates, when stocks have outperformed, bond yields tend to fall by less than average over the last few days of the month.   Similarly, when stocks have been crushed, yields tend to fall by more than average.

 
It may be the case that the fixed income sell off has exhausted itself, if only temporarily.   If so, however, the historical record suggests that it will not be month-end rebalancing that drives whatever correction we may see.
Previous
Next Post »

8 comments

Click here for comments
Anonymous
admin
November 28, 2016 at 2:44 PM ×

Not to forget a rerunof the previously rigged Austrian election on Sunday. The stamps stuck this time.

https://www.theguardian.com/world/2016/sep/12/austria-presidential-election-rerun-to-be-postponed-faulty-glue-ballot-papers

"A rerun of Austria’s presidential election has been postponed after the adhesive seals on postal votes were found to have come unstuck.

The rerun, which was ordered after complaints of anomalies in the counting of postal votes from the rightwing Freedom party (FPÖ), had been due to take place on 2 October. It will now be held on 4 December."

Reply
avatar
johno
admin
November 28, 2016 at 3:05 PM ×

Appreciate the post, MM.

My observation of the morning is: each time the Turkish 5Y bond has hit ~11% since the GFC, USDTRY has reversed. Curious. Of course, this time may be different, with the reflation/Trump and the eurodollar shortage and all.

Reply
avatar
Leftback
admin
November 28, 2016 at 4:14 PM ×

Thanks for the post. US fixed income is catching a bid today, perhaps the most obvious and important correlation in this context is CNY turning upward at last. The whole anti-dollar basket is firmer today, AUD, gold, GDX and JPY. Even TRY.

Like johno above, we're not trying to be too clever, our thesis is just this: DXY overbought and US jobs data are unlikely to reveal US "escape velocity", underlying growth rates of 4% etc.... let's face it, guys, we have been here many times before.

We fancy a dollar reversal so much here that we are long GLD and GDX calls. Regular readers will know we are no goldbugs at the Hammock, but we do like a power squeeze on an old-fashioned mean reversion trade.

Btw, for those following the bond bounce idea, the technical picture is much improved. The RSI and slow stochastic look good, all we need to confirm the bounce is a high volume buying day. Because of the likelihood of substantial short spec positions, the resulting squeeze might extend the move to surprising size. The long bond has come in below 3% today, and on a squeeze could even fill the Trump Gap down to 2.65%, providing substantial discomfort for those Bondmageddon fans who bet yields would break out to the up side, soaring with the eagle that is the US economy....

[LB is NOT Gary Shilling. Really.]

Reply
avatar
Anonymous
admin
November 28, 2016 at 5:03 PM ×

LB ... you have any views on REITs or preferreds these days?

Reply
avatar
johno
admin
November 28, 2016 at 5:12 PM ×

Italy referendum. Feels like "smart money" will buy a "no" outcome. Actually, I'd be more worried by a marginal "yes" outcome. A weak Renzi government is a government that could easily lose to M5S in the 2018 election, and with the constitutional changes, M5S could get decisive power in that scenario. Maybe BTPs are a buy on the "Renzi steps down" headline? I do wonder what that will do to the Monte recap, however. Already shaky.

Reply
avatar
abee crombie
admin
November 28, 2016 at 5:15 PM ×

Interesting study MM. if you wanted to be really clever i guess you could data mine the closing few minutes as well and see if there is some divergence.

speaking of 5yr, Johno, I think the US fiver is the tell for interest rates. While we seem to be holding the 3 year range at 1.8 (we did spike to 1.9), classic TA would argue for a break higher IMO, given the false breakout earlier in the year. But this is probably as good as any place for a cheeky long in rates

Reply
avatar
EschewObfuscation
admin
November 28, 2016 at 6:54 PM ×

A dumb question came to mind while reading comments on previous post:

Why would a company need to repatriate cash in order to use it for buybacks? Don't U.S. stocks trade overseas?

Reply
avatar
RH
admin
November 29, 2016 at 5:24 AM ×

One potential supporting fact-let for a bond bounce: the VIX has been higher at some point in December than it is today (13.x) every year since 2007 (earliest data I looked at). Ignoring the proximate GFC data: 4 of the last 5 years have seen it above 20 in Dec at some point and on the other time time it was over 15.

Kind of surprising and not really enough data for sane conclusions, but...

Reply
avatar