Deflationary Pressures Should Begin To Ebb
CNBC loves its “countdown” toy, and uses it whenever possible to give the time remaining before economic releases or market openings or closes. But when today it began using the clock to count down the moments until the signing of the health care bill it seemed oddly inappropriate. It was as if one could almost hear an oddly metallic female voice intoning pleasantly, as in a movie, “Your economy will self destruct in…ten minutes…and…fifty-two seconds…” The false precision in the countdown-meter, which included hundredths of a second for a bill signing ceremony whose timing is probably measured no more accurately than whole minutes, took on added ironic meaning when one considers the likely error of the many estimates of costs and revenues in the bill. I don’t think “close enough for government work” was ever meant to include hundredths of a second, and plus or minus 500 billion dollars for this landmark legislation would have to be considered a triumph of forecasting given the historical reliability of cost and revenue estimates.
It was yet another quiet day overall, although a 3-4 basis-point drop in swap spreads (probably due to the large amount of corporate issuance, which tends to pressure swap spreads when the issuer desires to translate the fixed coupon into a floating coupon) pushed the 10y swap spread to where only the 30y had previously gone: negative.
Some observers may get distressed by this: a LIBOR swap with a bank counterparty trades at a lower interest rate than 10-year government debt! Under the traditional explanation of swap spreads, which actually was true once, this would be cause for concern since if the swap spread is a credit spread, it implies that bank credit is better than sovereign credit. However, swaps now generally trade on a fully-collateralized basis, which means that the difference in credit is slight – the LIBOR rate is fully collateralized, while the Treasury rate is backed by the full faith and credit of the U.S. Treasury. Even if we still used the credit notion of a swap spread, it isn’t clear why negative spreads would be too alarming since after all, a collateralized obligation is generally better credit than an uncollateralized one for the same credit.
However, the conventional understanding of swap spreads observes that the main difference is in the floating rate. The receiver of the fixed rate on the swap also pays 3-month dollar LIBOR, which is a rate that does have some credit implications when things are really bad but most of the time is just a generic funding rate. On the other hand, the owner of a Treasury bond receives the fixed coupon but also has the benefit of being able to use the bond as high-quality collateral, thereby funding the security in the “repo” market at an attractive price. If the bond is “general collateral” (that is, it’s just any ol’ Treasury), the repo market will trade a few basis points below LIBOR, but Treasuries occasionally can go “special” and finance quite a bit below LIBOR.
So this is the main source of swap spreads: Treasuries finance at better rates than LIBOR most of the time, and sometimes they finance at rates much better than LIBOR. Except, that is, when interest rates are already at zero. Although technically negative repo rates are allowed, they are rare, so at low interest rates the possible advantage of the repo rate compared to LIBOR gets compressed. So while it is unusual to see negative swap spreads, it isn’t likely to signify anything about the quality of Treasury credit.
It does happen to be the case, by the way, that swap spreads tend to compress when Treasury supply increases. This happens because the chance of a Treasury security being scarce, and therefore going “special” in the repo market, is much lower when issue sizes are monthly at $15bln than when they are quarterly at $8bln.
Existing Home Sales came out today near consensus estimates, although as I noted yesterday the “consensus” is a term used loosely in this case. There isn’t very much in this report, especially given the recent noise, to get excited about but I thought I would give a longer-term perspective. I have been pointing out for a while that if you take Shelter out of core inflation measures, to get a look at what is happening to the non-bubble part of the economy, inflation has been accelerating for some time. This is of less concern if the deflation of the bubble is to continue for a while, but in many ways it looks like housing is approaching levels that reflect a more-normal historical relationship to the rest of the economy. For example, the chart below is a slightly-updated version of one that appeared in the December 2009 CFA Institute Conference Proceedings Quarterly in an article by Earl Webb entitled “Assessing Real Estate Markets: Pothole or Sinkhole?” It illustrates the bubble that the Fed was unable to discern.
The source for the median home price data is the National Association of Realtors’ Existing Home Sales report (and thus slightly differs from Mr. Webb’s chart in that he included all home sales, I believe); the source for median household income is the Census Bureau. The latter is reported with a one-year lag, so only 2008 is available, but we can make a reasonable guess and extend the chart to 2009. Note that all of these figures are in nominal, not real terms, but since inflation is operating on both values it shouldn’t have an impact on the ratio. I have also shown the long-term ex-bubble average.
The bottom line is that while housing might yet overcorrect, it looks like it is getting close to fair relative to incomes. That, in turn, means that unless incomes take a meaningful dip in real terms, the downward pressure from declining shelter costs are probably all in the pipeline by now, and the downward pressure on rents should gradually diminish over the next year.
If that is the case, then core inflation with shelter will over that period of time begin to converge with core inflation ex-shelter, and that implies much higher core readings a year or so from now.
With the exception of swap spreads, there was little excitement in the fixed-income markets today. The 10y note contract ended unchanged, with 10y yields at 3.58%. The S&P rose 0.7%, further confounding those of us who don’t see the nationalization of health care as a positive thing for the market (see my comment yesterday for a quick explanation of why it is very unlikely to be a net benefit to the market).
Tomorrow, the economic docket carries Durable Goods (Consensus: +0.6%/+0.6% ex-transportation following -0.6% ex-transportation last month) and New Home Sales (Consensus: 315K vs 309K last month). The 309,000 number was the lowest ever for New Home Sales, so the consensus call isn’t exactly looking for an aggressive bounce, but a further record would be somewhat distressing.
I doubt either of these figures will be market-moving. The best bet I think is for another fairly dull day.