CPI Number Tomorrow--What Bar Will You Celebrate In?
Are you guys excited?! Tomorrow the CPI number for March comes out. You may recall that Gary North has been strutting because of the flat February number, so I am very anxious to see what happens.
For those who can't stand the suspense, let me up the ante: The producer price index for March came out today, and was 1.1 percent higher than the previous month, i.e. it is running at over a 14% annualized rate. That certainly doesn't seem too deflationary to me.
For the real thinkers out there, who might wonder whether the PPI and CPI move together: The PPI is certainly more volatile, but they tend to move together. I hope this link works, but try checking out this graph. You see how the red line jumped up at the end? Which way do you guess the blue line will go? I will be very surprised if it's flat again.
Ah the joys of being a libertarian economist! You don't get the thrill of running people's lives, but it has its perks.
For those who can't stand the suspense, let me up the ante: The producer price index for March came out today, and was 1.1 percent higher than the previous month, i.e. it is running at over a 14% annualized rate. That certainly doesn't seem too deflationary to me.
For the real thinkers out there, who might wonder whether the PPI and CPI move together: The PPI is certainly more volatile, but they tend to move together. I hope this link works, but try checking out this graph. You see how the red line jumped up at the end? Which way do you guess the blue line will go? I will be very surprised if it's flat again.
Ah the joys of being a libertarian economist! You don't get the thrill of running people's lives, but it has its perks.
For the real thinkers out there, who might wonder whether the PPI and CPI move together: The PPI is certainly more volatile, but they tend to move together
ReplyDeleteBob,
What's with all this historical correlation business, of late, all about? Correlations in the past never mean they will continue into the future.
It is belief in correlations that killed Long Term Capital Management, caused the subprime crisis and will eventually carry out on a stretcher anybody who thinks that they can act on this type of historical data. Mises and Hayek are turning over in their graves.
Robert,
ReplyDeleteYour point is essentially correct, but ignores the Hayekian possibility of "pattern prediction." Yes, whatever correlations we may observe at any time between such macro variables may alter, but, during the time in which they hold, they permit a certain predictive power over the future. The trick -- which cannot be enabled by any statistical studies, however sophisticated -- is to judge when the pattern which held will break down.
Hey Annonymous,
ReplyDeleteHayek was not talking about "tricks" to judge when a data correlation pattern will break down.
Here's Hayek:
... we can never achieve a full explanation, or an exact prediction of the particular outcome of a given situation, but must instead be content with what I have occasionally called a "pattern prediction" or, earlier a "prediction of the principle." All we can achieve is to say what kinds of things will not happen and what sort of pattern the resulting situation will show, without being able to predict a particular outcome.
What Hayek is talking about here can best be understood by an example in meterology. We "know" that it will snow in the winter and not in the summer. This is the type of pattern prediction Hayek was referring to.
Nowhere close to a simple data correlation that Bob is trying to pull off when he writes:
For the real thinkers out there, who might wonder whether the PPI and CPI move together: The PPI is certainly more volatile, but they tend to move together. I hope this link works, but try checking out this graph. You see how the red line jumped up at the end? Which way do you guess the blue line will go? I will be very surprised if it's flat again.
Going back to the meterology example, data aren't collected (in the northern hemisphere) for everyday in February and a prediction is made that the snowiest day in February was the 12th, therefore, it is more likely to snow on the 12th. Hayek's "pattern prediction" is not about this type of data collection.
It is more like "the northern hemisphere gets less sunlight during December through February, thus it will tend to be the coldest and most likely time to see snow. However, there are too many variables to predict snow on any specific day. But the pattern will be snow in the winter months."
PPI and CPI moving together doesn't even fit basic Austrian business cycle theory. Deductively, money flows to capital goods sector first (PPI?), much later to the consumer sector (CPI?). There's a very loose fit, since the new money eventually works its way through the entire economy, but to try and make a prediction that CPI will be up this month because PPI was up, is nutty. (Especially given these are government manipulated numbers!!)
There's a lot of newly printed Fed money around so CPI may very well be up, but it will have little to do with the PPI being up this month, except in a very loose way.
"Red lines on a graph"? "Blue lines on a graph"? I stand by my statement,Hayek is spinning.
Wow, Robert, that quote is pretty far from making the case that Bob is not making a pattern prediction!
ReplyDeleteIn any case, what does it matter if Hayek meant this or not? Such predictions are possible, but when institional conditions change, the pattern breaks down.
Well, I'm glad you have decided to ditch Hayek in your defense ("what does it matter if Hayek meant this or not?"), since it is now clear, he was not theorizing about a Bob Murphy kind of prediction.
ReplyDeleteThese are examples of pattern predictions:
"The Federal Reserve has increased the money supply by over 7% during the last tweleve months, this is likely to result in a increased consumer price inflation in the future."
"The Federal Reserve has increased the money supply by over 7% during the last tweleve months. This has likely created malinvestments in the economy. If the Fed tightens, it is likely to lead to recession."
You write:
Such predictions [Murphy type] are possible, but when institional conditions change, the pattern breaks down.
I can't see how this kind of thinking means anything more than "The model works until it doesn't."
LTCM and the subprime paper buyers all thought the same.
Hayek's point, to bring him back into the picture, is that there are too many variables to know in advance what "institutional changes" could occur:
All we can achieve is to say what kinds of things will not happen and what sort of pattern the resulting situation will show, without being able to predict a particular outcome.
Bob is not only predicting an outcome, but a very specific outcome between March PPI and March CPI. Please explain the theoretical behind "if March PPI is up, March CPI has to be up".
Further, please tell, what "institutional changes" Bob should watch out for to know his model is breaking down, since "[s]uch predictions are possible"?
Mr. Wegner,
ReplyDeleteGary North said we are in a deflationary environment. He cited flat M1 growth.
I have pointed out that this is a poor argument.
North pointed to the flat CPI numbers in February as proof he is right.
Because North was so cocky about it, I have been anxiously awaiting the CPI number. The PPI number came out yesterday and was huge.
I'm not sure what you want me to do to keep Hayek and Mises from spinning in their graves. North made an empirical prediction that I think was going to be wrong.
Are you saying an Austrian economist can never make a falsifiable statement about economic matters, without being a heretic?
I always understood that the point was that theoretical ideas were not falsifiable except before other ideas. Austrian predictions are just as falsifiable as anyone else's.
ReplyDelete"Well, I'm glad you have decided to ditch Hayek in your defense ("what does it matter if Hayek meant this or not?"), since it is now clear, he was not theorizing about a Bob Murphy kind of prediction."
ReplyDeleteWow, again, Robert, since I said it wasn't at all clear, based on your quote, that he wasn't.
Annonymous,
ReplyDeletePlease forgive my error in misinterperting your reason for ditching Hayek. I was actually giving you credit for coming around to understanding of what Hayek meant by pattern predictions.
But, I now see you have not come that far.
So why did you ignore these questions posed to you:
Please explain the theoretical behind "if March PPI is up, March CPI has to be up".
Further, please tell, what "institutional changes" Bob should watch out for to know his model is breaking down, since "[s]uch predictions are possible"?
Brian N.
ReplyDeleteI always understood that the point was that theoretical ideas were not falsifiable except before other ideas.
If I theorize that inflation and unemployment can not both increase at the same time, and through empirical observation I see just that occurring, then empirical observation falsifies the theory.
Bob,
ReplyDeleteThere is nothing wrong with falsifications, as long as they are correct ones. But, the fact that March CPI is up doesn't falsify very much. Are you saying that becasue March CPI was up, we are in an inflationary period?
I think Gary North is very wrong in his deflationary forecast, but March CPI up doesn't prove that. A trend of generally higher CPI numbers over the next 12 months would falsify North's forecast.
Gary North is in error by jumping up and dwon about the February flat CPI number,just as you are to celebrate "in a bar" because March CPI is up. In fact, the only falsification that can be reached here is that you can't use one month's CPI number to forecast a trend, since you and North have come to complete opposite contradictory conclusions based on one months data.
Further, you do suggest that because March PPI was up that March CPI would be up. This is a completely falsifiable proposition, since the very chart you use in your post shows many instances of PPI being up in a given month when CPI was down.
There'a loose connection between PPI and CPI because they are both influenced by money printing, but to suggest this includes a trend at the month to month level is simply false. Look at that chart again!!
Mr. Wegner,
ReplyDeleteOK, generally speaking, I have to say that I am not nearly as dumb as you apparently think. Some clarifications here:
There is nothing wrong with falsifications, as long as they are correct ones.
I wasn't defending "falsifications," I was defending the assertion of a falsifiable statement. A purely Misesian statement on economic theory is not falsifiable. So since I took you to be criticizing me on those grounds, I was asking you if an Austrian economist could ever make a falsifiable statement, such as, "I betch the CPI number will be positive tomorrow."
But, the fact that March CPI is up doesn't falsify very much.
I agree North's views might not be construed as "much," but they certainly make him look bad. I'm dying to see if he comes back to this topic again, and how he will spin it.
Are you saying that becasue March CPI was up, we are in an inflationary period?
It depends what we mean by "inflationary" of course. But no, I was saying it means North is wrong for claiming that the CPI numbers prove he is right and everyone else is wrong.
I think Gary North is very wrong in his deflationary forecast, but March CPI up doesn't prove that.
It does if he wants to say that flat February numbers prove he is right.
Gary North is in error by jumping up and dwon about the February flat CPI number,just as you are to celebrate "in a bar" because March CPI is up.
Oh wow. That was a joke based on the fact that it's dorky. I was making it sound as if there were a big NCAA game or something, and people would "ring it in" at a bar. The celebration was for the number's release, not that it would be high.
Further, you do suggest that because March PPI was up that March CPI would be up. This is a completely falsifiable proposition, since the very chart you use in your post shows many instances of PPI being up in a given month when CPI was down.
Are you kidding me? If I say, "Gunshots to the head are usually followed by blood loss," are you going to chastise me for a falsified proposition? I mean, if you shoot a corpse blood won't shoot out. I need to go read Mises again.
Bob,
ReplyDeleteA purely Misesian statement on economic theory is not falsifiable.
Well this is a tricky statement, since you have to define what you mean by a "Misesian statement". You are almost saying if someting is true, it is true (and therefore non-falsifaible). Since it is true, how could it possibly be falsifiable? On the other hand, is it fair to say that Gary North is a Misesian and he holds a theory that we are entering a period of deflation? If there is nothing but inflation for the next 12 months, can we not say that North's in-correct Misesian theory has been empirically falsified? Or do we say it is not Misesian, which gets back to the impossibilty of falsifying a Misesian statement because by definition it is true.
Oh wow. That was a joke based on the fact that it's dorky. I was making it sound as if there were a big NCAA game or something, and people would "ring it in" at a bar. The celebration was for the number's release, not that it would be high.
I think you made a wise career choice, economics over comedy.
You then quote me:
Further, you do suggest that because March PPI was up that March CPI would be up. This is a completely falsifiable proposition, since the very chart you use in your post shows many instances of PPI being up in a given month when CPI was down.
You then say:
Are you kidding me? If I say, "Gunshots to the head are usually followed by blood loss," are you going to chastise me for a falsified proposition? I mean, if you shoot a corpse blood won't shoot out.
I think part of the problem here is that I am coming somewhat from a traders perspective. If you tell them that March PPI was up and that therefore March CPI should be strong. There are a bunch of traders that are going to jump on the proposition. Further they are going to start programming computers to alert them to jumps in PPI, so that they can trade ahead of an up CPI.
You have to be real careful, exact and accurate when talking to traders, they really take throw off claims seriously. No I don't think you are "dumb", but I do know that when you say something in a light matter a trader could look at it differently.
I remember once giving a 20 minute speech, where I threw out some kind of a line about copper. One line, that went something like this "..and metals like copper, silver, iron, etc."
Three months later I ran into a trader who was at the speech. First thing out of his mouth, "Hey are you still bullish on copper?"
Given the Long Term Capital Management blow up and the subprime blow up, both of which I believe occurred, in part, beacuse of poor economic methodology (mostly bad historical correlations), if we are not going to be sticklers for proper, exact methodology in every uttrance, who is? How will they ever learn if there isn't someone standing up for exactness in an area where the failure of exactness has caused enormous damage to the economy?