We have run some numbers using the New York Fed's Primary Dealer Statistical database, which in combination with the Fed's POMO operations in the second half of September, result in some peculiar conclusions. First, we have summarized all of the Fed's coupon POMOs from September 15 through the end of the month (we excluded the $550MM TIPS POMO from September 28): as the table below demonstrates, in five operations Brian Sack repurchased a total of $16.4 billion in Treasurys. So far so good. Yet when juxtaposing these presumed repurchases of coupon securities, with the disclosed holdings of coupon Treasurys by Primary Dealers, something does not jive: to wit - PDs disclosed a total of $6.8 billion in coupon (not Bill) USTs held as of September 15 (split between holdings of 1-3 Year, 3-6 Year, 6-11 Year, and 11+ Year positions). One would surmise that courtesy of over $16 billion in Coupon (not Bill - this is important) monetizations, PD holdings would at least declined, if not by the fully monetized amount, then at least partially. No such luck: in fact UST holdings increased by $400 million headed into September 29, or the last end of quarter number. What did decline, however, and to a much greater extent, was the PDs holdings of Bill securities, which dropped by a 2010 record of $50+ billion over the same period. And it did not stop there: adding in changes in the other four PD disclosed security holding categories, Agency Coupons, Discount Notes, MBS and Corporate Bonds, and the total decline since September 15 was a whopping $62 billion in PD holdings! So while POMO was used by the PDs for everything but what it was intended to monetize, the same primary dealers who were the benefactors of the Fed's guaranteed UST bid instead used the end of quarter, FRBNY-facilitated window dressing to not only not offload coupons, but to dump everything else, and use the proceeds to buy stocks, thereby explaining both the massive ramp into the end of September, and also the ongoing attempt to flush NYSE shorts, which as of September 15, were still near 2010 records.
The table below demonstrates the divergence between what PD UST holdings should have done and did do, despite $16.4 billion in FRBNY coupon monetizations.
So what does this mean: simple. PDs kept all Coupon Treasury asset classes flat, while dropping Bills holdings by $50 billion in two weeks, in exchange for investing said capital for risk assets. In fact, on September 29, total PD holdings (Bills and Coupons) dropped to just $6.3 billion, or the lowest level since February 24, when they were negative. This can be seen on the chart below.
To be sure, all of this is part and parcel of the ongoing end of quarter window dressing effort, which long after Repo 105 has been uncovered, and despite the SEC's inquiries into such practices by all banks continues. Nowhere is it more visible than on the charts below: the first one shows total PD weekly holdings since 2005.
As is now completely expected, the September asset plunge brought the total PD assets to within a hair of the lowest aggregated holding total seen during Q3, and $64 billion away from the total recorded just two weeks earlier! The chart below shows the min, max and EOQ PD positions during any given quarter. It is all too obvious that for the 6th quarter in a row, Primary Dealers closed their quarters with a flush of all assets, in hopes of closing at or very near lows recorded during the quarter.
As noted above, the primary culprit for the plunge in assets has to do with the liquidation of Bills and near-dated coupons. This is visible on the chart below which shows the weekly change in the five UST holding classes demonstrated by PDs. The summed two-week plunge in Bills is the biggest on record, and at $50+ billion, it is even greater than the $41 billion in Bill exodus recorded in the week following the Lehman crash, and the second highest Bill outflow recorded for $33 billion in June 2009.
We have no doubt that once the week of October 5 data is disclosed by the FRBNY, that the Bill holdings by PDs will surge, now that the latest window dressing period is over. The problem is that this will have required a major sell off in risk assets. Which begs the question: if indeed PDs saw their non-stock assets surge (and we will know this for sure shortly), and will retail investors continuing to withdraw money from the market, and with the Fed's POMO actions having no impact on the one class (coupon USTs) it is supposed to be influencing, just who is doing all the stock buying out there? Of course, that is a very rhetorical question. It is now obvious that the PDs wanted nothing more than to cause the short-covering spree in the second half of September to become self-sustaining. As we demonstrated earlier, the NYSE SI was at near 2010 highs despite the massive rally in the first half of September. We will soon know if this number has indeed declined after the ongoing blood feud the Fed and the PDs have with all shorts. Should this number not have declined, expect some wild activity from the PD-Fed contingent as the two throw every kitchen sink they can get their hands on in a last ditch attempt to stimulate the short covering rally. Already, as we pointed out last week, rumors of SPY and IWM buy-ins were proven correct, after the market surged by almost 2% the day following the rumor. Very soon State Street will be forced to recall all shorts, as any selling or shorting activity become virtually illegal. At that point, it will be time to go all in. However, not the stock market, but gold, which is where all the stock pessimists have moved to: after all, let the Fed have its Dow 36,000. If it is achieved at the expense of gold $36,000 pretty much everyone will be content.