|GDP Report (again!)
||[Dec. 5th, 2013|11:44 am]
So, the first revision of the advance estimate for 3rd quarter GDP came out.
I had commented before on the advance estimate.
And, like the BEA, I am updating and revising my entry there.
(1) Real GDP grew at a 3.6% pace, a substantial upward revision from the 2.8% - if I may be allowed to give myself a pat on the back, I had previously said "I don't put much stock in advance estimates". This, my friends, is why. I'm also going to make a prediction for the final revision which will come out in a month: I think it will be below 3.6%, but above 2.8% - it's pretty typical for the final revision to get a result that is between the first two estimates.
(2) I had misinterpreted the data before. The 1.8% increase in prices was an annualized rate - that number's revision kept it the same as before.
Based on this, I would have revise some of my predictions.
At this point, there is no evidence that the excess reserves are actually hitting the economy - as a result, I don't see any reason that a crash would be imminent at this point. The evidence for an artificial boom is not clear at this point, so I don't know that an inevitable crash is coming. However, I still worry about excess reserves, as those have the potential to create a boom-bust cycle if they hit the economy in any significant way.
||[Nov. 7th, 2013|03:19 pm]
So, the Bureau of Economic Analysis just put out it's advance estimate of third quarter GDP. Looks like real GDP increased at a 2.8% annual pace, a bit higher than the 2nd quarter's 2.5%.
Since macroeconomists are supposed to have a view of this, I suppose I will.
(1) I don't put much stock in advance estimates. They get revised twice before we have the final number - and revisions can be large (sometimes a full percentage point or more).
(2) The increase in the growth rate was driven by imports slowing, an acceleration in private inventory investment, and an acceleration in state and local government spending.
(3) The price index for gross domestic purchases went up by 1.8% in the third quarter - that's 7.2% at an annualized pace, if I'm reading the data right.
(4) The areas of big growth were primarily in investment - residential and nonresidential structures coming in in first and second place.
I find the combination of data somewhat worrisome. Austrian Business Cycle Theory suggests that periods of artificially low interest rates result in unsustainable booms - typically investment driven - that will inevitably crash when interest rates rise. Why does this make me worried?
(1) Interest rates are obviously very low right now. Whether they are "artificially" low is a fair question. A good way to get at that: is investment outpacing savings? That's not entirely clear - recent data actually suggests that gross saving is actually higher than investment - so rates might be low "for a good reason". At the same time, the massive increase in the monetary base makes it fee unlikely that the low interest rates aren't the result of Federal Reserve policy - that is that rates are "artificially low". The argument goes two ways: one says "The Fed has increased the monetary base - therefore rates are artificially low." The other - which I'm coming more around to - is "The Fed has increased the monetary base - but the money is sitting in banks rather than entering the economy - so the Fed's actions haven't necessarily affected interest rates in any substantial way." We could expect this continue as long as banks don't think there are good opportunities out there. But, here the increase in private inventory investment suggests things may be changing in that regard. One interpretation of inventories says "Oh noes! We're making more stuff than we're selling! We're going to have to cut back production in the future!" It ends up that this interpretation is basically wrong. In reality, most of the variation in inventories isn't from finished goods that just don't have a buyer. It's from partially completed goods that just weren't finished at the time the data was collected. That being the case, an increase in inventories suggests that businesses are optimistic enough about the future that they're starting more production processes. I worry that it's only a matter of time before this increased optimism leads to accelerated lending by banks... and that time might actually have already passed...
(2) Now, we come to the problem with the "money is just sitting in banks" argument - the price index for gross domestic purchases went up very fast in the 3rd quarter - this suggests that the money the Fed created IS entering the economy and being spent - driving prices up - which there's reasonable evidence for. (Though I have to note that the inflation rate as measured by more conventional measures - like CPI - remains pretty tame.) This will create pressure on the Fed to increase interest rates...
(3) Which would lead to the crash. (Actually, it would lead to a kind of crash even if the interest rate wasn't artificially low to start with.)
(4) How bad would the crash be? Well, that depends on a few things. Consider two different crashes. The first was the "tech bubble" which crashed in 2000-2001. In this case, the tech bubble was funded largely by equity, and the real investments were largely in things that had good alternative uses (computers, knowledge, etc.). The crash happened, and we recovered fairly quickly. The second was the "housing bubble" which crashed in 2007-2009. This was funded largely by debt, and the real investments were largely in things that didn't have good alternative uses (houses). So, the transition has been difficult. So, what's happening now? The current investments are in things that probably don't have good alternative uses (structures), but they seem to be funded largely through equity - at least, borrowing has been small (home mortgages are declining on net, though there has been a somewhat troubling increase in the amount of corporate debt being issued).
So, in my view, odds are good there's a crash of some kind coming - but I'm thinking it probably won't be as bad as the housing crash proved to be, but would be worse than the tech bubble popping was.
But, all of this is just based on advance estimates.
|Fed Chairs and Recessions...
||[Nov. 6th, 2013|02:56 pm]
So, it was observed - partially in jest, I think - that Janet Yellen should steer clear of the Fed chair job because it's a trap. But, it raised a real question: is there a connection between changing of the Fed chair and recessions setting in? Here's what I found...
At this point, we have had 14 different Fed chairs since 1914. Of those, 4 did not experience a business cycle "peak" (that is, a turning point from the economy booming to falling into a recession/depression). 10 of them DID experience some kind of peak. What was the timing? (One of them took over in the same month as the economy peaked in 1923 - ouch. For him, I'll just use the NEXT peak which didn't happen for about 3 years after that.)
For those that experienced a business cycle peak, there was an average of 26 months between when they assumed the Fed chair and when the peak happened - so just over 2 years between taking over and the economy taking a turn for the worse, on average.
But, we can't read too much into this. "On average" is an important phrase - and the economy is often not "average". (For example, while the average is 26 months between taking the Fed chairmanship and the next peak, NO Fed chair saw that exact number of months pass.)
So, doing some absolutely atrocious statistics (because it's likely to be pretty inaccurate, not because it's difficult...), and assuming a normal distribution (probably shouldn't, but oh well!)...
Given that Janet Yellen is going to take over in February 2014, and assuming that she won't be one of the lucky chairs that doesn't experience a recession at all, there's a roughly 68% chance we'll have a business cycle peak sometime between March 2015 and May 2017, and a roughly 95% chance we'll have a business cycle peak between February 2014 and June 2018. Which... isn't very informative since the bands are so wide.
To get at the significance of the new chair, though - let's suppose that business cycle peaks are purely random. If so, then there being 33 months where business cycle peaks happen in about 1872 months (since June 1857) suggests there's a 1.75% chance that any particular month will be the peak of a business cycle - and a 98.25% chance it won't. Assuming independence (a HORRIBLE assumption here in all likelihood), the odds of there being at least one business cycle peak in a 26 month period is about 37% - so the new chair does increase the odds somewhat in the March 2015 to May 2017 period (from 37% to 68%). The odds of there being at least one business cycle peak in a 52 month period is about 61.5% - a new chair drives that up to 95%.
Now, we want to be careful about cause and effect. It's perfectly possible there's no causal connection - there's a dreaded "3rd thing" that leads to new Fed chairs AND business cycle peaks having this relationship. But, there are two possible direct causal connections I can think of: Either the new Fed chair causes the ensuing recession - or foreseeing the coming recession leads the previous chair to leave. Interestingly, most estimates suggest that monetary policy takes about 2 years to "work" - so it could easily be either...
This, friends, is the joy of uncertainty.
|Concluding Thoughts on the Government Shutdown
||[Nov. 6th, 2013|12:02 pm]
I know, I know. I should have gotten over it by now - the shutdown ended a while ago. But, I've been intending to write this up since then and just hadn't made the time.
So, here we go:
Point #1: I've becoming convinced that the Democrats were the ones who were truly to blame for the shutdown.
I know, this goes against the conventional view. But, it's the result of asking a very simple question: "Who benefits?" If you see an undesirable situation, it happened for a reason. If it happened because people did something you're going to find the people responsible by asking the question about who benefits. Who benefits from the shutdown? The Democrats do - and did. It covered up the disastrous roll-out of the Health Insurance Exchange, and it made the Republicans look bad. So, either the Republicans are to blame and are just idiots, or the Democrats are to blame and played the situation very well. I'm more inclined to believe the latter (though you're free to believe the first).
There will be a simple test to get at whether I'm right. The Democrats had been claiming that they're willing to negotiate some of the provisions of Obamacare, but not in a hostage situation (the shut down was apparently a hostage situation...). If I'm right, they're going to give almost no ground on Obamacare. Though they might, given the technical problems, provide some delay in the individual mandate - but I expect that delay to be minimal.
Point #2: Debt-ceiling crises are easy to prevent.
Simply pass this law: "Whereas the debt ceiling serves no practical purpose because it can be raised by the people incurring the debt at their own whims, it is abolished." Easy.
Point #3: Budget debates are also easy to prevent.
Why does the government have a budget anyway? No, I'm serious. It makes sense to have a budget if (1) you have an income that is mostly fixed by some outside source, and (2) you actually stick to it. Neither of these is true for the government. It can raise more money if it wants to simply by introducing new or higher taxes. (The Laffer curve being a limitation here, of course.) And its budget is more or less meaningless when it comes to restraining spending. So, why not adopt a more piece-meal funding solution? Simply pass a general rule that resembles the old PAYGO system - every bill is required to be self-funding either through raising additional tax revenue or through making spending cuts elsewhere. In fact, we have such a bill in place right now (though it has a distressing number of exemptions from the rule). That being the case, what purpose does the budget serve? It's not at all obvious.
I very much look forward to watching how #1 plays out...
|More thoughts on Government Shutdown
||[Oct. 7th, 2013|04:46 pm]
So, the shutdown continues. And my life continues to be mostly unaffected. Two exceptions: (1) the BEA shut down its website, which is inconvenient for research and teaching purposes. (2) I'm having to ignore more of my Twitter Timeline (and, to a lesser degree, Facebook News Feed) because it has turned into political drivel that really shouldn't be compressed to the appropriate lengths for those media. (Don't many of us complain about politics being reduced to slogans? Yet, we still feel fine making political tweets that are limited to 140 characters...)
As far as I'm concerned, I think my original conclusion still stands: the Tea Party is not some "sliver" of the Republican Party that everyone else hates. It's a sizeable force that can actually change policy.
I'd add to it my wife's cynical conclusion as well: the Democrats are going to let the shut down continue as long as they can, because the Republicans are taking the blame for it. (I was convinced by this article, which shows how the Democrats plus a small handful of Republicans could force through a clean CR without Boehner or the Tea Party folks having to give any ground whatsoever.)
One thing I find remarkably distressing, though, are the awful analogies that I'm seeing Democrats throw around. Let me offer two:
( Don't let Pelosi do hostage negotiations or fight fires...Collapse )
The really, truly fascinating thing is that no one is asking the big question: Why aren't politicians who have to win popular votes to get/keep their jobs bothering to do things that are popular? 60% of Americans say it's more important to end the shutdown than to modify Obamacare. Why aren't Republicans doing that? Nearly 60% of Americans want to see changes in Obamacare. Why aren't the Democrats allowing for a 1 year delay in the individual mandate to hammer out something more popular? (Being the latest Republican offer I've heard of.)
The answer. (Hat-tip to @sahilkapur on Twitter) The members of the parties each support their own party - and don't want it to compromise.
( Why this is a problem, but doesn't mean we're more divided...Collapse )
This raises all kinds of big questions, apart from the simple, most pressing, one ("How can we get the government shut down dealt with?").
How can we restructure our system to avoid such partisanship in the future? (Given the reality that the "Republic" form of our government has become largely fictional, would we be better off with proportional representation and coalition-building like much of Europe does?)
Will (or can) a third "moderate" party arise? (Independents are a larger group than either Republicans or Democrats.)
Personally, I doubt that either of these will happen - but it's not entirely clear how else (apart from secession) the deep divisions in the US can be resolved.
|More Crude Empirical Testing...
||[Sep. 11th, 2013|02:56 pm]
So, I have some weird results. The summary: I can't find a theory of interest rates that actually "works".
Liquidity preference: claims that increases in money supply and decreases in money demand lead to lower interest rates.
Test result: Predictions based on this are right 21.43%, and move the wrong direction 53.6% of the time. (No movement in interest rates the otehr times.)
Time preference without money: claims that increases in the Consumption/Investment ratio reflect increased time preference which drives up interest rates (and vice versa).
Test result: Predictions of interest rates based on C/I ratio are right 24.5% of the time - more the wrong direction 42.4% of the time. (No movement in interest rates the other times.)
Time preference with money (a position I've advocated before): claims that C/I ratio reflects "demand" for present goods in some sense, but this can be offset by increases in money supply which are a present good. So, decreased C/I ratio PLUS increased money supply results in lower interest rates. (or vice versa) Other combinations are uncertain.
Test result: Predictions of interest rates based on the combined movement of C/I ratios and Money supply are right 24.6% of the time, and move the wrong direction 39.3% of the time.
End result: all three theories - applied in this very crude form - are missing something - and a big enough "thing" that they are wrong more often than right.
Alternative theory: interest rates are exogenously determined by policy in our system, and everything else responds to that...
C/I should be predicted to increase when interest rates fall. (We save/invest more at that time.)
Money demand should be predicted to increase when interest rates fall. (May as well hold money longer.)
Theory of exogenous interest rates...
Predicts direction of C/I changes correctly 68.6% of the time - incorrectly 31.4% of the time.
Predicts direction of Md changes correctly 78.8% of the time - incorrectly 21.1% of the time.
Conclusion: interest rates explain a lot very well but are very difficult to explain.
|Liquidity Preference Theory of Interest: a Crude Empirical Test
||[Sep. 11th, 2013|12:23 pm]
One of the highlights of Keynes's General Theory was his "Liquidity Preference Theory of Interest". More or less, Keynes argued that it's the supply of and demand for money that determine interest rates.
So, I devised a crude test. I looked at M2 money supply and M2 velocity (which is inversely related to money demand), and looked at the impact on the prime rate.
To narrow it down, I looked for times when M2 and its velocity were both increasing - suggesting that the money supply was rising while money demand was falling - by Keynes's theory, this should mean we'd get falling interest rates.
From 1980 until the present, there have been 56 months in which the conditions of rising supply and falling demand were satisfied - but only in 12 of those did the prime interest rate fall. In 30 of them, the prime interest rate actually rose - the exact opposite of what the theory would have suggested.
I feel like there's a paper here somewhere...
|Open Letter to College Students
||[Sep. 9th, 2013|03:10 pm]
Dear College Students,
Welcome! This may be your first year on campus - this may be your last. But, as you spend whatever time you do with us, I'd like to offer you some friendly advice from a member of the faculty.
( Cut for length...Collapse )
Wishing you all the best,
|Metro Unemployment Rates (Or: When Summarized Statistics are Misleading)
||[Aug. 28th, 2013|02:38 pm]
So, the BLS just put out the new Metropolitan Area Employment and Unemployment Report (reporting for July). It looks good! About 90 percent of Metro areas saw an improvement in their unemployment rate.
So, I got curious. How is the Canton-Massillon area doing?
On the face of it, it looks good! The unemployment rate fell from 7.5% to 7.3% - a noticeable (if small) improvement.
But, then I look deeper. Why? Because of math and the way these statistics are calculated.
( Educational stuff!Collapse )
So, it's true that the number of unemployed fell by 500 people. But, 80% of that can be explained by the fact that 400 people left the labor force (or, possibly, moved out of the area). Only 100 found jobs.
To put it another way: let's say that those 400 kept looking for work rather than leaving the labor force. Based on employment changes alone, the unemployment rate would still have rounded to 7.5%, not 7.3%.
Ohio as a whole sees a similar - but not quite as bad - story. the number of unemployed in the state of Ohio fell by 11,700 - but 5,000 (just under 50%) of those can be explained from people leaving the labor force. (Some cities have very nice numbers - the nearby Cleveland-Elyria-Mentor had 14,400 JOIN the labor force, and unemployment fell by 3,700 people - suggesting 18,100 people found work who didn't have work before.)
( Clearing up a confusionCollapse )