Do Deficits Matter?

|
 |  Includes: SPY, TENZ, TIPZ
by: Michael Ashton

According to Bloomberg, Treasuries on Friday fell and stocks rose “as U.S. Economic Data Top Forecasts.” Stocks added 0.6%, while 10-year note yields perked up to 3.32%.

The economic data that beat forecasts was the trade deficit, which was mildly better than expected, and the Michigan Confidence figure, which at 74.2 managed to beat the 72.5 expectations, but didn’t even reach a new high for the year. It should be noted that the low for the Michigan number in the last recession was 77.6.

Oh, and by the way, the budget deficit for November was 150.4 billion, more than $12 billion worse than expected and the biggest November deficit ever. To the extent that these second-tier numbers are better than expected, do you think that Federal largesse has anything to do with it? I’d be more impressed if the deficit was improving as well.

Moreover, China overnight increased bank reserve requirements for the third time in the last few months. After the prior two hikes, markets sold off. This time, there was nary a ripple – perhaps because investors figure growth is so robust we no longer need to rely on China as an engine of growth?

Overlooked too, it seemed to me, was the fact that the leaders of Germany (Chancellor Merkel) and France (President Sarkozy) met today and jointly announced that they reject the notion of the European Union issuing bonds – which of course would mainly have value because of the economic power of Germany and France – and are opposed to expanding the €440 billion rescue fund that was set up in May – and which costs fall predominantly on Germany and France. To be clear, Merkel and Sarkozy reiterated that they completely support the common currency, but they also made clear with their actions that they are more willing to let periphery nations flounder alone than to let the problems of the PIIGS besmirch the good names of Germany and France.

So stocks rallied, according to the soothsayers, because economic data topped forecasts. Given the actual distribution of news today, though, this seems to me to be a hopeful justification. Stocks look to me like they are rising now because they are at new post-Lehman (OTC:LEHMQ) highs, and some investors are afraid of missing the boat. And bonds are falling because it is December, and they already have fallen quite a ways, and in the bond world you generally don’t want to fade moves made in December no matter what the cause because of the size of the short-gamma hedging need compared to the available liquidity.

If the news is so good, then let’s see how the stock market holds up next week when it gets real data. Retail Sales is on Tuesday along with PPI and a Fed meeting; CPI, Empire Manufacturing, and Industrial Production are on Wednesday; and Housing Starts, Initial Claims, and the Philly Fed report on Thursday. I don’t think that we know anything more today about the economy, with Michigan and Trade in-hand (especially given China and the budget), than we knew yesterday. But by next Friday, we will actually have some real information.

I will leave you this week with one more fun chart. Today’s deficit figures put the running 12-month total at a mere $1.288 trillion. Assuming we have roughly 2% growth this quarter, this means the deficit is around 8.7% of GDP.

Now, we all know that big deficits should lead to higher real interest rates, right? Well, take a look at the following chart, which shows the 10-year TIPS yield against the rolling 12-month deficit as a proportion of GDP. What we see is rather the opposite – real interest rates have instead declined as the deficit has increased.

(Click to enlarge)

Note: Last 2 points of Surplus/Deficit line are estimated based on 2% growth rate in current quarter.

Let this be a lesson against the careless use of statistical inference! What is happening here is that both deficits and interest rates respond to a common factor, and that is economic growth. The first period of deficit deterioration, which largely overlaps the first big decline in real yields, is associated with the recession of the early 2000s. Deficits improved and bond yields rose during the tepid expansion of the middle part of the decade, and then the second recession of the decade sent both deficit and yields lower. So real yields are not responding to the deficit, nor the deficit to real yields; they are both responding to rotten economic conditions (and where we are on the chart should give you some pause about expecting robust growth ahead).

However, look instead at a scatter plot of TIPS yields as a function of the deficit. The relationship is pretty good, but interestingly it is not linear. I’ve plotted two regression lines here: one is linear and the other is polynomial of order 2. The polynomial curve fits better (notice the higher R2), which confirms what our eyes seem to tell us when we look at it.

(Click to enlarge)

Note: Same data as above, but arranged in scatterplot form.

In other words, the indirect relationship (through growth) isn’t clean. Real yields are not as low as they should be given growth that is bad enough to lead to a deficit as large as it is. This either means that (a) we are getting bad mileage out of our deficit spending, and running one too large given the underlying economic growth, or (b) real yields are higher than you would think they should be given how punk growth is. I suspect both of these are true, but the implication of the latter is more worrisome because we can always improve the profile of what we’re wasting money on.

Higher real yields suggests that the government is having to pay a bit more to fund its deficit (in real terms), even though the Fed is buying scads of the debt. Now, if we want to be generous we can suppose this is because private borrowers are competing for the capital more than they were, but look again at the first chart and see where the inflection really happened. This is a phenomenon associated with the huge deficits resulting from the latest recession.

Indeed, my regression lines are probably too generous. The clump of dots on the right appears to be roughly linear, but then the clump of dots on the left represents total outliers. If the linear-vs-nonlinear approach is “right,” then real yields are about 75bps higher than we would expect them to be otherwise but that’s just the point estimate and you can plausibly argue the effect isn’t large. But if the dots from the most-recent period are best viewed as outliers relative to the “normal” relationship from the last recession and the expansion periods around it (basically, one full cycle since TIPS were first issued), then real yields are a couple of hundred basis points too high, or the deficit is over 5% larger than it should be given the underlying growth dynamic, or some combination of the two. Any way you slice it, I think the first chart (the “happy” one) is misleading, and the scatterplot (the “unhappy” chart) is more illuminating. So, do deficits matter? You tell me.

Disclosure: No positions