More Disturbing Inconsistencies

If we compare consumer spending on services with hours worked in the service sector employment report, we find a huge surge in implied service sector productivity. We suspect the service sector consumer spending series is overstated, which means real GDP growth, while already horrible, is also overstated. As we've mentioned before, government surveys of service sector activity are not as timely or as complete as those available for the manufacturing sector, so a lot of trend extrapolation goes on in the Washington, D.C., data mills, at least on the initial estimates of service sector activity.

Since productivity growth is used to estimate unit labor cost growth, costs also must be understated for the service sector as well, which further implies service sector profits are overstated. This should all come out in the wash in future revisions, but for the meantime, it is very likely that the GDP, consumer spending and profit realities are worse than currently reported.

Transfer payments have reached a higher share of personal income than interest and dividend income. Together, that means over 30% of the U.S. personal income flows are now being earned for no increase in work devoted to production of goods and services (although we do recognize loans earning interest may, in fact, finance increased production, rather than financial market speculation). The more people get paid without directly producing goods and services, the higher the inflationary bias likely to arise in an economy. Purchasing power without production is another, perhaps more accurate, way of depicting the old saw about "too much money chasing too few goods."

Money doesn't chase anything. People spending money who did not produce anything to get the money - now, that's a recipe for inflation (or less disinflation/less deflation, than otherwise would be the case from favorable cost and supply conditions, to put the point more generally). In this regard, the major rise in transfer payments from 1966-76 may have played a role in the original onset of the stagflationary '70s, while the surge in interest income from 1978-82 may have contributed to the second wave of stagflation. Higher money incomes without increased production tends to lead to higher prices, unless, of course, saving rates among money income recipients rise sufficiently. Treasury Inflation-Protected Securities have been one way to hedge for an eventual inflation result, although they are no longer as cheap - but then again, neither are most inflation hedges, which have, for the most, part been bid up year to date.

We have been investigating which sectors have been accumulating Treasuries lately, as we believe placing new Treasury bond issuance may become more challenging not just because of the huge supply forthcoming, but because the quantitative easing measures adopted by the Fed and other central banks are designed in part to force investors out of the risk spectrum and away from cash and default-free Treasury holdings. Results from the Fed show broker-dealers, foreign investors and money market mutual funds have been the largest buyers of Treasuries of late. We are especially concerned the large buildup of long Treasury positions at primary dealers is unlikely to be sustained and there may be a rout in the Treasury market reminiscent of 1994, when Goldman Sachs had to borrow from the Japanese in order to stay afloat after getting caught the wrong way round in yield curve carry trades.

With 10-year Treasuries hitting 3.36% and passing through their 200-day moving average this week, our concern is looking less and less theoretical. We asked seasoned bond veteran Lou Crandall at Wrightson Associates about the unusual buildup in dealer position in Treasuries, and this was his assessment, which we have corroborated elsewhere:

"The large structural short positions that dealers developed in the middle part of this decade were a function of their market-making and hedging strategies. As a general matter, dealers would hold inventories of less-liquid spread products such as mortgages, ABSs and corporates on the long side, and then hedge them by shorting Treasuries of comparable duration.

"Once the liquidity crisis hit in the summer of 2007, a couple of things happened. First, the basis risk on those trades exploded, because spreads became highly volatile. Treasuries were no longer as efficient a hedge for private instruments as they had been. After Bear Stearns, you also started to see a rapid contraction in dealer balance sheets - even if Treasuries had still been an effective hedge for private-sector instruments, dealers were trying to lighten their inventories of those private obligations.

"In 2009, a couple of additional factors have come into play. Dealers wanted to front-run the Fed's purchases of Treasuries, which probably led to a modest amount of speculative long positions in Treasuries. More importantly, the Treasury market in May has adopted new delivery protocols that impose a 300-basis point penalty on anyone who fails to delivery a Treasury on time. The prospect of that fee created a lot of uncertainty in the market in April, and led some dealers to conclude that it would be safer to conduct their market-making activity from the long side, rather than the short side, of the market in the near term. (Just about everyone came to that conclusion about the bill sector, but some applied the same logic to coupons.)

"The increase in outright long dealer positions in Treasuries is close to having run its course. We could continue to see increases for a few more weeks as dealers complete the transition. However, we are seeing a restructuring of dealers' market-making business model, rather than an increase in outright long-term holdings in Treasuries driven by an investment view on the part of the dealers."

Essentially, dealers covered their spread trades by purchasing Treasuries and selling agency mortgage-backed securities and corporate bond positions. Some of them reconfigured themselves as banks, where procedures for the risk weighting of capital favors holding Treasuries. These, along with some procedural changes, are essentially one-off transitions. We, therefore, do not believe broker-dealers are likely to be big buyers in future Treasury auctions. Unless U.S. macro data start to get more alarming soon - and auto production cuts could dampen the Institute for Supply Managements' numbers, while the minor consumer bounce in Q1 was really mostly over in January and could cool further in Q2 - Treasury yields are likely to surge higher. So far, the backup in Treasury yields has not taken 30-year fixed mortgage rates higher, but this is the type of challenging set-up in the Treasury market that could squash attempts to stabilize the U.S. housing market. If you were planning on refinancing your mortgage this year, you may wish to consider acting sooner rather than later.