Bob Murphy on Efficient Markets
Bob Murphy is frustrated, but he’s lashing out at the wrong theory:
I understand the Efficient Markets Hypothesis, and I think it’s a very good way to take a first crack at the markets. The thing that annoys me about many EMH proponents is that they think they are being empirical and scientific, when they often are clearly able to explain any outcome in their framework. Steady growth? Just what EMH predicts. Massive crash? Just what EMH predicts. In practice, the EMH is non-falsifiable, which is ironically the criticism many of its proponents level at others.
The EMH is most certainly “falsifiable.” It’s been tested in many ways. Some people even claim that it has been falsified, although I’m not convinced. In the tests that I think are the most relevant the EMH comes out ahead. (Stocks respond immediately to news, stocks follow roughly a random walk, indexed funds outperformed managed funds, excess returns are not serially correlated, or not enough to profit from, etc., etc.) I assure you that if stocks responded to news with a 12-hour lag, or were clearly far from being a random walk, even Fama would reject the EMH. BTW, is ABCT refutable? If so, how? (I don’t regard refutability as the most important test of a theory–usefulness and coherence are better tests in many fields.)
I think this following passage from Scott is a tad slippery:
Murphy seems to suggest that the fact that Austrian economists were not surprised by the volatility is a point in their favor. But why? Who was surprised? If you had asked me a year ago “Do you expect occasional volatility, up and down?” I would have said yes, and also that I had no idea when that volatility would occur, or in which direction the market would move.
Look, there was nonstop coverage of this on NPR when it happened. They were trotting out all kinds of people, including Austen Goolsbee, to make sure Americans kept their money in Wall Street. I’m not making this up, give me a break.
Monday showed the biggest intraday point swing in history. (Granted, you would want to look at percentage swing for a better comparison, but I can’t find such a ranking.)
And according to this guy’s analysis, by one measure of market volatility-the VVIX-Monday blew the previous record out of the water:
Bob should not rely on NPR for his stock analysis. And you really should look at percentages, not absolute changes. He talks like the US just experienced a stock market crash, at a time when the market less than 10% below its all time high! Monday was not a particularly big deal in terms of stock market history. Yes, the volatility was unusually large by the standards of 2015, but this has been an unusually placid year. Back in 2008 we had weeks and weeks of non-stop action that was roughly as volatile as Monday. I’d guess that one could find many hundreds of days throughout stock market history where the stock market fell by as much as it did on Monday.
But let’s say I’m completely wrong, and stocks were historically volatile. Even then Bob’s wrong, as he completely misunderstood my quote. By “more of the same” I merely meant the most likely outcome for the market was what we have observed in the past. There are periods of stability and periods of volatility. If the S&P is at 2108, and someone asks me where I expect it to be in two weeks, I’d say 2108, or maybe 2109 (to reflect a gradual upward trend.) But that does NOT mean I actually expect the stock market to equal precisely 2109 in two weeks time. Rather it reflects the fact that I view it as being equally likely to rise or fall. I actually think it far more likely that it would be considerably higher or lower than at that specific point estimate. Bob’s frustrated that I’m not making market forecasts that can be refuted, but that’s because I don’t think it’s possible to forecast the market.
And I feel the same about the business cycle. The US has recessions every 5 or 10 years, China less often. We don’t know when the next one is coming. Usually I’m right, because any given year I say that growth is more likely than a recession, and when the recession actually occurs then I’m wrong. I was wrong in 2008, as I did not predict a recession in 2007.
Now if last year Bob had said there’d be a crash on August 24, 2015 and it happened, then more power to him. But as far as I can tell he simply posts “I told you so” blog posts after every minor pull back, before the market again soars to new heights. Then another modest pullback, and another “I told you so.” I honestly don’t know what we are to make of all that. Does the Austrian model provide some key to predicting the stock market? If not, why talk about stock moves as if they support the model?
If you want to say I’m a broken clock, or that we should wait and see what things look like in three years, etc., that’s fine. But come on, don’t act like predicting “more of the same” two weeks ago is consistent with what just happened.
Bob doesn’t tell his readers that the link is to me discussing the Chinese business cycle, not the US stock market. Some might even say that’s misleading, as his post implies it applies to the US stock market. But I won’t complain; I stand by my previous “more of the same” as being a wise prediction, and I’ll apply it to the US, to China, to stocks or business cycles. Whatever Bob wants. And if I visit the Sands casino and I predict that the little bouncing ball will land on a red or black, and it ends up on the green 0 or 00, I’ll stand by my prediction that red or black were the most likely outcomes. That I made a wise prediction.
In most areas of life we judge competence by track record. Doctors, lawyers, engineers, etc. But that doesn’t work for market forecasters. Track record tells us nothing about the competence of stock pickers. It doesn’t tell us whether their future predictions will be better than those with a poor record. And I think that really frustrates people. It goes against common sense than past performance is not an indicator of competence. But it just isn’t. I think that might be why Bob is exasperated by my placid agnosticism.
PS. Obviously I do know that volatility is serially correlated. I hope readers don’t think I’m THAT stupid. Thus it goes without saying that for the Chinese market a “more of the same” prediction two weeks ago implies a prediction of continued high levels of volatility. Which is what happened. You may disagree with me, but please don’t assume I’m a complete moron.
Tags:
28. August 2015 at 04:56
@And if I visit the Sands casino and I predict that the little bouncing ball will land on a red or black, and it ends up on the green 0 or 00, I’ll stand by my prediction that red or black were the most likely outcomes. That I made a wise prediction.
Ah, that’s a perfect analogy that almost nobody, even in finance, understands. They will continue to gloat ad nauseam that you were wrong, pointing to 0/00, and not having a clue re how foolish they look.
Thanks for staying with the science!
28. August 2015 at 05:15
“The EMH is true/false” has always struck me as unhelpful.
How efficient? That is the question. And the answer, which I think is empirical, is “pretty damn efficient”. Warren Buffett might say “not so efficient” but I don’t have his eyes or brain.
“Track record tells us nothing about the competence of stock pickers.”
Well, a lousy track record tells me something. But I think your point here is that, even over periods of 10 years or longer, the sample size is not sufficient to reliably distinguish luck from skill. Very frustrating.
In my mind, this is testament to the efficiency of yet another market – the market for stock pickers. Tough racket. Unloved bunch.
If the whole world decided tomorrow to put their money into Vanguard index funds, one stock picker could make easy money. In a sense, the stock pickers are the guys on the front lines, duking it out and doing the analysis and price discovery which the Vanguard S&P 500 index then parasitizes to exist.
I’m not advocating ‘active investing’ for 99% of us. From the outside, stock pickers look hapless, but on the inside, it’s a shark tank. Stick to your day job, people.
But they are part of a marvelous (yet demonized) financial infrastructure that allows something like the Vanguard S&P 500 Index to exist, a vehicle for allowing people to reliably transfer consumption over time in an historically unprecedented and efficient manner.
But most people don’t understand the stock market, and they fear it. In the land of New Normal 3%-4% interest rates, this fear will be expensive.
28. August 2015 at 05:24
Am I wrong to think that these big swings in asset prices can be irrational, but that this is not really a point against EMH, which remains roughly true?
28. August 2015 at 05:37
I posted this over at Murphy’s site, but I think it may have been moderated out of existence. Not sure.
//Comment begin:
Two things:
1. Taking the S&P500 Since 1960 or so, there have been about 100 days with price changes as large or larger than the one we experienced on Monday – so it happens, say, twice a year. Rare, but not unprecedented.
2. The VVIX that you [Murphy] reference is not a good measure of stock market volatility – it is instead a measure of the volatility of the volatility of the market. The VIX is a better measure of market volatility (implied from the options market again), and this measure shows that Monday was a volatile event, but not historically unprecedented – roughly on par with the events in summer of 2011, but still, even if we use the VVIX, there have been 7 such readings (as your post shows) since 2007, so again, maybe a once per year thing.
Consequently, I would say there is validity to the opinion that there was nothing particularly special about the market volatility we’ve experienced over the past week or so, but I won’t comment on whether or not it was surprising. I think the definition of a drawdown like that shows that some people were surprised.
//Comment End
And also, yes, volatility is highly serially correlated, with obvious implications for forecasting. This is well known.
28. August 2015 at 05:43
@Michael Byrnes, I think it’s better to think of these things as reflecting times of heightened uncertainty.
Binary good/bad outcomes replace the smooth continuum of expected outcomes the market envisions in less uncertain times.
It’s hard to look at the stock market in 2008/2009 as lurching back and forth between two discrete potential futures.
Between 9/29/2008 and 3/23/2009, the S&P 500 recorded 9 of the worst 20 days seen in the last 65 years.
But here’s the funny thing: during that same period, the market also saw 10 of the top-20 one day gainers since 1950, including the biggest one-day advance in the last 65 years (up 11.61% on 10/13/2008).
The 1987 crash shows a similar pattern of ‘market in the process of deciding between two divergent paths’. Between 10/19/87 and 1/8/88, the market had 3 of its worst 20 days since 1950, but also 3 of its best 20 days.
28. August 2015 at 05:53
Bob should not rely on NPR for his stock analysis.
Obligatory XKCD
But seriously, I think that what a lot of critics miss is that there is as much wisdom in saying “I don’t know” (e.g. EMH) as there is in making a correct prediction; and the closer your process is to a mere coin flip the more wise the former and the less the latter.
So the real problem is that people like Bob Murphy think that the market is not like a coin flip. But your assertion would be that if it isn’t, then make a correct prediction. Then they would answer back, just because Bob Murphy can’t predict the stock market doesn’t mean that the underlying process is not causal, it just means that it is too complex for Bob Murphy to predict.
To which your answer is, if the market is so complex that it is indistinguishable from a coin flip, then that means that you can’t pinpoint the causal mechanism. Which is a real bummer for those who have “the theory” that purports to explain everything.
28. August 2015 at 06:22
One thing I noticed about ABC guys like Bob Murphy is that they seem to look at daily gyrations in the stock market as indicators of the overall economy. That and they are always thinking the market is over valued and set for a devastating crash. This crash never really comes though. The closest we’ve seen in our lifetime was 2008 but look we are back. But they think we should see a massive contraction to bring us back to 1980 numbers or something. It is weird and there is no getting through to these people because they won’t accept evidence because they have this doomsday feeling in their blood. It is like those Christians always looking for the rapture.
28. August 2015 at 06:32
Lots of goods comments, I have nothing to add.
28. August 2015 at 06:37
‘…the EMH comes out ahead….’
Which is all you need to know. It works better to explain the facts than any other theory.
28. August 2015 at 07:19
Seems to like Bob was so anxious to finally be able to say he was right that he jumped the gun badly. I think people like Bob, Ron Paul, Peter Schiff need an 75% or greater drop to really be right but they lose patience. I think Bob was expecting the fall to continue but he just did not have the patience to wait and see.
BTW: I am betting heavily against that kind of a fall. I do not see any way outside of nuclear war for a fall like that to happen. Even if they where right about a coming inflation, I think that stocks would hold most of the value.
I think that stocks are still undervalued compared to any other investment. You can buy 3% dividends in fairly safe low debt companies with historys of increase divided faster than inflation. I am not even convinced that Government bonds are safer than stocks of companies with low debt that pay dividends. Conditions can cause an erosion of a T-bill’s value. The 10 t-bill rate is 2.16%, KO pays 3.38% and is diversified across the globe. A lot has to happen for the T-Bill to do better than a diversified group of companies similar to KO?
28. August 2015 at 07:32
‘In a sense, the stock pickers are the guys on the front lines, duking it out and doing the analysis and price discovery which the Vanguard S&P 500 index then parasitizes to exist.’
…
That’s a great point. At the same time, I wonder if the indexes and shadow index funds (and most managed funds must own certain percentages of each sector) aren’t the ones setting the prices, as they are making most of the trades.
28. August 2015 at 07:34
@Floccina
Indeed I think you are correct. The S&P 500 is already back to above it’s level on Friday’s close, and has retraced more than half of the total drawdown.
Not making any statement about where it goes from here, of course.
28. August 2015 at 07:34
Both the EMH and ABCT pale in explanatory power to my theory:
‘Goddidit’
Which is all you need to know. It works better to explain the facts than any other theory.
(Don’t believe in any gods? Then try ‘shithappens’. Either way, no need for any more curiosity because I just finished science!)
28. August 2015 at 07:36
Brad DeLong: “A Cautionary History of US Monetary Tightening”
https://www.project-syndicate.org/commentary/fed-monetary-policy-tightening-risks-by-j–bradford-delong-2015-08
28. August 2015 at 07:46
DeLong makes the same mistakes that Prof Sumner makes: Overly loose money creates problems, the central bank tries to head off the problem, chaos ensues because there is no pretty way to deal with bad debt.
The implicit suggestion in DeLong’s view, seems to be that the way to deal with a problem like overextended S&L’s or overextended mortage industry would be to make credit even easier. But that’s the road to permanent zero rates, permanent bailouts for bad lenders and distorted market prices. And with so much lending going for foolish purposes, you wind up with stagnation.
28. August 2015 at 07:50
Great stuff 4:40 into this video!
http://www.bloomberg.com/news/videos/2015-08-27/china-must-free-float-yuan-in-order-to-cut-rates-woo
28. August 2015 at 08:01
Sumner: “Bob’s frustrated that I’m not making market forecasts that can be refuted, but that’s because I don’t think it’s possible to forecast the market.” –
Yet Sumner does feel that you can target NGDP by printing money. Strange, he doesn’t make the connection that people holding onto their money or spending it quicker (velocity) is not a constant, which will make a mockery of targeting NGDP, akin to pushing on a string at times.
Sumner does not understand that humans are unpredictable.
28. August 2015 at 08:38
Ray, it’s just a sample size of one, but you are utterly predictable.
28. August 2015 at 08:38
There are a number of novel properties of volatility of equity markets: it 1) clusters, 2) mean-reverts, 3) generally increases more when the market declines than when it rises.
28. August 2015 at 09:11
Also relevant – a good post by Wesley Gray @ Alpha Architect:
http://blog.alphaarchitect.com/2015/08/28/daily-stock-market-return-perspective-71926-to-62015/
28. August 2015 at 09:40
ssumner writes:
Stocks respond immediately to news, stocks follow roughly a random walk, indexed funds outperformed managed funds, excess returns are not serially correlated, or not enough to profit from, etc., etc.) I assure you that if stocks responded to news with a 12-hour lag, or were clearly far from being a random walk, even Fama would reject the EMH.
From empirical observation:
If investors aren’t expecting market moving information, it them anywhere from two to four minutes to respond to the new information in a way that moves stock prices. (Example: The Bernanke Noise Event).
If investors are expecting such information, or are using automation technology programmed to respond to potentially market moving news events in case they occur, the response is nearly instantaneous (Example: The Flash Crash of 23 April 2013, or the announcement of previously-embargoed economic data).
If one looks at just stock prices in a vacuum, their volatility appears to follow a random walk, but a better explanation is that they are more prone to follow a Levy flight, which looks like a random walk that is periodically punctuated by large movements. Or in other words, stock prices appear to move according to a random walk until they don’t.
The assumption of randomness breaks down however when we factor dividends and the future expectations associated with them into the picture. If their variation were really random, the variation of stock prices about a central trend, when one exists, would look like this, and not like this (the latter fails several of the old Western Electric tests for the appearance of unnatural patterns in data that’s expected to follow a Gaussian, or normal, distribution with no autocorrelation.)
28. August 2015 at 10:19
Brian, You stole my line!!
John and Ironman, You really need to think about the real shocks to the economy that cause stock prices to change. When you do it’s obvious that you wouldn’t expect a normal distribution, and that you would expect big drops to occur more often than big increases, and also clusters of volatility, even if the market were efficient.
28. August 2015 at 10:23
Scott:
How about if you expand the scope of the test? Maybe you can’t predict the rise and fall of the stock market, or time a recession, but do you believe you can predict the rise and fall of national economies over longer time horizons based on structural factors such as access to resources, good feedback mechanisms(e.g. free-markets), defensible borders, sound political and legal institutions, government debt levels, work ethic and so forth.
28. August 2015 at 10:26
The most obvious refutation of the efficient markets hypothesis is algorithmic trading. If you can train a computer program to execute trades on your behalf and consistently make a profit, even if it is only a few cents per transaction, then that would unequivocally show that there are inefficiencies in the market. As such trading algorithms do exist, the market is less than 100% efficient.
Yet, I still believe that the EMH is broadly true. Markets do quickly incorporate new information, and the market is difficult to predict. Market volatility and market crashes do not refute the EMH. They reinforce it. Unpredictable volatility is the very nature of efficient markets.
28. August 2015 at 10:34
Markets are clearly moderately efficient.
Some structural factors and psychological biases effect the market.
The market was at 2000 on friday afternoon. That lead to option guys having to sell stock to hedge changes in their put exposure. Which then lead the market lower and forced leveraged guys into margin calls. Is that efficient? Or a structural factor overriding efficiency?
Chinese market is clearly less efficient than the US market. Their are fewer investor bases to even things out and Chinese stock punters are inexperienced.
Agree that ETFs are basically parasites. They work because they don’t have to pay for stock analysts who are making stock markets more efficient. They do pay a bit for this on timing as their buying/selling can be partially predicted and front run.
28. August 2015 at 11:32
With nearly 20 years of following the “market” the most valuable lesson I will share with my descendants is this: Prognostications of the stock market are for entertainment purposes only, no matter who is giving them.
That said, Austrians are the most entertaining prognosticators. They are also the most often wrong.
BTW, Schumpeter was an Austrian, but does he count as an “Austrian”? There seems to be much that he “got right” about Capitalism, how it works and where it leads. But I suppose he got the timing wrong.
28. August 2015 at 11:52
Part of the trouble here is with the very term ‘prediction’. I think we should agree that this term applies only to *confident* and *precise* statements about the future, but Bob Murphy seems to have in mind only non-probabilistic statements, such as, “The S&P 500 index will be between X and Y one year from today,” whereas it should apply also to statements of the following kind: “The probability is 95% that the S&P 500 index will be between X and Y one year from today.” You are willing to make statements of the latter kind, and so you do, in a sense, “predict” the stock market, though not with Murphy-type absolute (non-probabilistic) statements. And if you are doing it as well as it can be done, *you will be wrong (in the sense that in a year the index will fall outside your limits) 5% of the time*. When this happens it will not, in itself, *refute your model*, nor does its failure to do so show that your model is empty or useless. Probabilistic statements are valuable, too, though irrefutable by single outcomes.
And isn’t it uncharitable to interpret *anyone’s* predictive statements as absolute rather than (at least implicitly) probabilistic? The whole Popperian scenario of absolute predictions *refuted* by outcomes is unrealistically oversimplified.
28. August 2015 at 12:02
“We don’t know when the next one is coming.”
-Given that there has been no expansion of greater than ten years in recorded U.S. history, and money is unlikely to start to become very tight until 2016-2017, I’m guessing 2018-2019.
“But it just isn’t.”
-How do you know that? Is there some kind of study on this?
Also, what’s the tiny smiley face at the very bottom of the blog? It always struck me as weird.
28. August 2015 at 12:15
“That said, Austrians are the most entertaining prognosticators. They are also the most often wrong.”
-Do you have any study on this?
28. August 2015 at 12:28
@Harding, Each person must judge what is entertaining. As for accuracy I know of very few market prognosticators who warn of “easy” money and who have provided timely market advice. Being wrong 9 times out of 10 means one is not providing very useful advice.
The problem with market pundits crying about “loose money” is that while such a policy will eventually dislocate the economy (which is my complaint about NGDPLT) in the meantime it benefits stocks. So if one thinks money is loose one ought to be buying stocks!
28. August 2015 at 13:09
@Ironman
IIRC, there are versions of the martingale representation thm for plenty of Levy processes with jumps. So the whole financial math edifice around FTAP basically still works even without Brownian motion.
But EMH doesn’t depend on a particular Markovian process to be true, in fact, prices will follow a process that is a function of information entering the system, which is a function of actual events occurring in the real world. So EMH could hold even if your model for stock prices is terribly naive.
28. August 2015 at 13:43
Scott, I was just expounding on your PS about the nature of volatility. I don’t think I was intending any other point than that.
In my work, I focus on statistics. I consider the normality assumption something that can be convenient computationally or analytically, but I don’t have any issue discarding it when I need to.
Moreover, it’s important to note that a statistical process can be unconditionally non-normal, even if it is conditionally normal.
Nevertheless, I think your analysis of real shocks and equity shocks is more true conceptually than statistically. It can be hard to tease out the impact of real variables like industrial production on equity markets.
28. August 2015 at 13:59
Carl, Yes, I predicted the rise of China back in 1980.
Doug, Well if the EMH is 99% true, then that makes it one of the truist model in all of the social sciences, doesn’t it?
E. I just observe that economists can’t predict them. So if a model exists, it’s top secret.
John, I agree. I look for natural experiments–stock market responses to news.
28. August 2015 at 14:50
@ssumner
-Doesn’t that just tell us most economists are incompetent?
28. August 2015 at 14:59
Okay. If national economies are to some extent predictable then I would think the markets within them have to be somewhat predictable: an index fund in a stock market in structurally sound Country A should eventually outperform a similar fund in structurally unsound Country B. In other words, I may not be able to out-invest an index fund in country A’s stock market, but I can out-invest the average index fund in Countries A and B simply by choosing an index fund in Country A.
How does the EHM define my act of choosing which national market to invest in? Is it a market activity or is it considered a non-market activity?
28. August 2015 at 15:25
@Philo, quantum mechanics is probabilistic (at least when it comes time to make a measurement). Yet Popperian ideas work OK there AFAIK.
28. August 2015 at 15:39
Bob should not rely on NPR for his stock analysis. — Scott Sumner
I have skimmed all the comments here and (unless I missed it) not *one* of you guys bothered to point out to Scott how ridiculous that particular jab was? There are evidentiary rules in place, to protect even those defendants that you are sure are guilty.
Remember kids, when Scott is trying to convince you his opponent is wrong, he will only be as intellectually honest as you force him to be.
28. August 2015 at 15:55
Bob, Njnnja did at least mention it:
http://www.themoneyillusion.com/?p=30337&cpage=1#comment-398384
…maybe not in the way you would’ve liked.
28. August 2015 at 16:20
Excellent blogging.
By the way, Wall Street has gobs of money and operates in a free market, in terms of hiring talent.
Surely, they would hire Bob Murphy if he could make useful predictions.
28. August 2015 at 16:27
“Carl, Yes, I predicted the rise of China back in 1980.”
-Pics or it didn’t happen.
28. August 2015 at 16:46
I predicted the Chinese stock market crash back in June when I criticized a story published on Bloomberg that heralded the PBoC for ‘lowering rates to clean up a debt mess.’ Where’s my 15 minutes of fame?
I am half kidding of course, what I actually said was it was more likely tight money – and stocks usually reflect it quickly. It’s not that hard to spot.
28. August 2015 at 19:56
Sumner wrote;
“The EMH is most certainly “falsifiable.” It’s been tested in many ways. Some people even claim that it has been falsified, although I’m not convinced.”
That’s precisely because you use EMH in a non-falsifiable manner.
What you are really saying there is just “Some other folks believe EMH is falsifiable, so that is enough to refute Bob’s argument that one of his pet peeves is how so many EMH proponents believe they are being empirical when they are really not.”
Let that one sink in.
What Bob is saying is that DESPITE the claims from EMH proponents that EMH is falsifiable, the way many are actually going about their thoughts and writings, they are clearly using it in a non-falsifiable manner.
You merely pay lip service to EMH being falsifiable, but the evidence and theory against EMH is rejected, not because the lack of theory and evidence, but because you treat EMH as a priori.
In other words, what Bob is saying is that your insistence and belief that EMH is falsifiable, contradicts what is going into your usage of it when comparing and contrasting various theories.
You say EMH predicts both no abnormal returns and abnormal returns.
You say EMH predicts both booms and busts taking place.
You say all this because you believe EMH is true regardless of the data.
Now no doubt right now you might be thinking of a set of events that in your mind would falsify EMH. you might even offer them as what would convince you EMH is falsified. Yet even those events you have already chalked up to pure luck. “Just wait a few more years and Warren Buffet will suffer a recession!” And if it doesn’t happen in the next few years, then you say it all over again.
Sound familiar?
28. August 2015 at 20:25
[…] in this post Scott first lectures me on how the EMH is falsifiable. Right, I know the academic EMH in the […]
28. August 2015 at 20:29
Hey Bob, if we wanted to talk bad about Scott we could go over to your blog and do it! We may not all agree with Scott but the heure d’oeuvres at his parties are excellent.
Scott, are familiar with the work of Vernon Smith on experimental economics? You can look up his Nobel Prize lecture titled “CONSTRUCTIVIST AND ECOLOGICAL RATIONALITY IN ECONOMICS” On the premise of what Smith discusses I ask you: Are the governments of the world rational in not embracing the theory of NGDPLT? If so what are they afraid of that would justify their reservation? If governments of the world are acting irrationally, why do you think this is?
29. August 2015 at 01:33
If recessions are caused by declines in NGDP and if we believe that NGDP futures markets are good at predicting changes in NGDP, then doesn’t that mean that recessions are predictable, at least by NGDP futures markets participants? I can understand why EMH implies that it should be difficult *to profit* from knowledge of an impending recession, but I don’t see why recessions should necessarily be unpredictable.
Conversely, if recessions are unpredictable, then does that imply that central banks always or almost always do a good job: they rarely have reason to believe that their policies will do anything other than to maintain stable NGDP? For example, when the ECB raised rates in 2011, was it predictable or unpredictable that such policy would cause a double-dip recession. If unpredictable, then was the rate increase the right (a priori) policy, the policy that would make future NGDP just as likely to be above target as below target, given information available at the time?
29. August 2015 at 02:59
“If the S&P is at 2108, and someone asks me where I expect it to be in two weeks, I’d say 2108, or maybe 2109 (to reflect a gradual upward trend.) But that does NOT mean I actually expect the stock market to equal precisely 2109 in two weeks time. Rather it reflects the fact that I view it as being equally likely to rise or fall. I actually think it far more likely that it would be considerably higher or lower than at that specific point estimate.”
Scott, you are 110% correct.
29. August 2015 at 03:03
…and for the record, I’m not a big fan of EMH, just not worth arguing over as it is good enough, longer time frames makes it even better.
29. August 2015 at 03:28
“But I think your point here is that, even over periods of 10 years or longer, the sample size is not sufficient to reliably distinguish luck from skill. Very frustrating.”
For active traders, just change the granularity to monthly, weekly, or daily.
29. August 2015 at 05:31
E. Harding, If you could predict the business cycle, then yes it would.
Carl, No, the EMH says the fact that a country is faster growing is already priced into asset prices—perhaps that’s why China has high P/Es.
Bob, That was a joke.
Bonnie, Well done.
Dan, NGDPLT is a policy, not a theory. And what makes you think they are not embracing the policy? Central banks change slowly. First the profession changes its views, then central banks gradually convert. It happened with inflation targeting. And now the profession is moving one by one to NGDPLT. We are winning, if you haven’t noticed.
BC, That’s actually a very good question, and the answer is very complicated. My view is that a NGDP futures market would not have been able to predict the 2008 recession until is was underway, but none the less would have been able to prevent it, or at least make it milder. Here’s a possible sequence:
Late 2007, NGDP futures market would have predicted slowdown in 2008, but no recession. Had we been following NGDPLT, then monetary policy would have been made more expansionary at that time.
Then the slowdown occurred. During the second half of the year, an NGDP futures market would have been predicting that the slowdown would turn into an outright recession. Notice that this is a successful prediction of a worsening of the recession, but it doesn’t count as predicting a recession because the recession actually began in December 2007.
Here’s a driving analogy. I can predict when the steering of the bus will lead it to drift off the center of the lane, but not when the bus will veer off the road and crash. If I nudge the bus back to the center each time it begins to veer off, then the bus never crashes.
An NGDP futures market cannot predict when the central bank will take the necessary nudges to keep NGDP track, but it can actually help the central bank do those nudges if it wants to.
29. August 2015 at 05:32
BC, I’ll do a post at Econlog.
29. August 2015 at 06:26
“The EMH is most certainly “falsifiable.” It’s been tested in many ways. Some people even claim that it has been falsified, although I’m not convinced. In the tests that I think are the most relevant the EMH comes out ahead. (Stocks respond immediately to news, stocks follow roughly a random walk, indexed funds outperformed managed funds, excess returns are not serially correlated, or not enough to profit from, etc., etc.)”
Yes, EMH is falsifiable. It’s also a perfect example of the “blackboard theory” Buchanan warned about. ABCT suggests that sometimes markets don’t accurately incorporate information from credit markets, which are themselves distorted by CB policy. Part of the distortion in asset markets is due to the difficulty in determining whether a change in the interest rate reflects credit market fundamentals or CB distortions. That’s all ABCT really says about asset markets, IMO.
“BTW, is ABCT refutable? If so, how? (I don’t regard refutability as the most important test of a theory-usefulness and coherence are better tests in many fields.)”
If you’d read Roger Koppl’s comment on David Henderson’s post, you wouldn’t have had to ask this.
http://econlog.econlib.org/archives/2015/08/robert_p_murphy_1.html#346547
Ultimately I’m not defending Bob’s piece. I think it was a misapplication of the theory. That makes me even more frustrated to see the flurry of posts telling everyone how absurd ABCT is based on Bob’s article.
29. August 2015 at 06:40
Benny Lava, you wrote:
“One thing I noticed about ABC guys like Bob Murphy is that they seem to look at daily gyrations in the stock market as indicators of the overall economy. That and they are always thinking the market is over valued and set for a devastating crash. This crash never really comes though. The closest we’ve seen in our lifetime was 2008 but look we are back. But they think we should see a massive contraction to bring us back to 1980 numbers or something. It is weird and there is no getting through to these people because they won’t accept evidence because they have this doomsday feeling in their blood. It is like those Christians always looking for the rapture.”
One thing I noticed about EMH guys like Benny Lava is that they seem to look at thousand year period gyrations in the stock market as the only indicators of the overall economy. That and they are always thinking the market is under valued and set for a permanent skyrocketing towards infinity. This launch never really comes though. The closest we’ve seen in our lifetime was 2007 but look we are back. But they think we should see a massive expansion to bring us back to Weimar numbers or something. It is weird and there is no getting through to these people because they won’t accept evidence because they have this faith based devotion to socialist money and duty bound feeling in their blood. It is like those Milleniallists always looking for the second coming.
29. August 2015 at 07:05
Levi, You said:
“That makes me even more frustrated to see the flurry of posts telling everyone how absurd ABCT is based on Bob’s article.”
I hope you are not referring to this post, as I don’t base my views on ABCT on Bob’s posts.
I base it on Austrians who claim money was easy in the early 2000s because interest rates were low. Or that Austrian economists can somehow see foolish malinvestment occurring that markets are unable to see. That sort of thing.
I’m a market monetarist. So far the Austrians have offered me absolutely NOTHING that would make me want to drop my model and adopt theirs. If they have some good arguments, I’d love to see them. I’m open to new ideas. Yet every time an Austrian says “you should read this paper” it offers me absolutely nothing of interest.
29. August 2015 at 09:36
Scott,
Sorry, when I said “Bob’s piece” I was referring to the original article that set all this conversation in motion.
I’m sure we have different standards of proof and I’m by no means a “representative” of the Austrian school, but if loan growth is strong, bond purchases are strong, and rates are low, that sounds like a rightward shift in the supply of credit (high quantity, low price). I haven’t looked at the data, but that’s how I think about credit markets right now. I’m a micro guy, so of course I may be missing something important.
I know you don’t like getting paper suggestions, but since you brought it up, here are a couple you may not have seen.
Here’s a working paper which analyzes papers published since 2008 the financial crisis that discuss Austrian theory. The abstract explains the paper well so I’ll just give you the link.
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2363560
Claudio Borio’s 2012 BIS working paper takes an empirical look at the financial cycle with an eye toward Austrian theory.
http://www.bis.org/publ/work395.htm
29. August 2015 at 12:03
Perma bear Taleb’s hedge fund just made a billion dollars last week:
http://www.wsj.com/articles/nassim-talebs-black-swan-fund-made-1-billion-this-week-1440793953
30. August 2015 at 04:49
Levi, Sure that might be easy credit, but easy credit has nothing to do with easy money–they are completely unrelated concepts.
I agree that at the time 2008 looked superficially Austrian, to people who don’t understand monetary economics. But subsequent events have not been kind to the Austrians. I’d say the same about the 1930s. In 1930 the Austrian explanation was riding high—by 1933 it was pretty much discredited.
30. August 2015 at 11:54
“Sure that might be easy credit, but easy credit has nothing to do with easy money-they are completely unrelated concepts.”
So do you actually know which is relevant to ABCT? Are the two policies “completely unrelated” in the real world?
“I agree that at the time 2008 looked superficially Austrian, to people who don’t understand monetary economics.”
Are you sure that your understanding of ABCT is sophisticated enough to accurately make that judgment? Nearly every one of your posts is about NGDPLT, so I have some understanding of where you’re coming from on that, but lots of bloggers have leveled specious criticisms (most recently that I’m aware of, Noah Smith) of ABCT, to the point that they were laughable to those of us who have read professional work on ABCT.
30. August 2015 at 13:13
Sunner wrote:
“Levi, Sure that might be easy credit, but easy credit has nothing to do with easy money-they are completely unrelated concepts.”
This comment shows Sumner has insufficient understanding of both monetary economics and banking/financial economics.
Easy money defined as the Federal Reserve System (government plus member banks) inflating the money supply is almost always brought about by easy credit.
The money the Fed creates via OMOs does not directly increase the individual bank balances of the entire population, out of which “the money supply” aggregates like M2 gets its name. It only increases the bank balances of those who own the member banks and thus the member bank’s cash accounts, of whom the Fed solely deals with.
No, the way individual bank balances of the whole population are increased in the aggregate is by way of bank credit expansion. That is how new, additional cash accounts are increased. Banks issue new loans, and the depositing of those loans in a bank then serves as an increase in the quantity of money for the entire population together.
As such, easy money not only has “something” to do with easy credit, in our monetary system easy money depends almost exclusively on easy credit.
If the Fed issued billions and billions of new currency by way of OMOs, but the banks did not expand their lending such that no new bank balances are created other than those created by the OMOs themselves, then aggregate spending would not increase much at all.
Sumner has the false belief that the only way new money is created is by way of Fed OMOs. That the base money supply is expanded by the Fed, and for some unexplained reason, this base money supply immediately becomes the property of everyone with chequing accounts, such that NGDP rises when that base money supply is subsequently spent on goods.
In reality, the easier credit expansion is, the higher the aggregate money supply will become, and the higher NGDP will become, which is precisely what Sumner defines as easy money!
To get a data driven sense of just how off Sumner is in his false claim that easy money has nothing to do with easy credit, just look at a chart that includes base money supply versus total aggregate money supply such as M2 (or M3 if you go back before the Fed stopped reporting it).
You will see that the base money supply makes up only a fraction of total money supply. The rest is ALL credit expansion driven. In other words, when NGDP expands, this is the Fed depending on member banks to expand more credit in order for aggregate spending made by the whole population to go up.
“I agree that at the time 2008 looked superficially Austrian, to people who don’t understand monetary economics. But subsequent events have not been kind to the Austrians. I’d say the same about the 1930s. In 1930 the Austrian explanation was riding high””by 1933 it was pretty much discredited.”
No empirical events refute Austrian theory, because Austrian theory is a priori. It is a theory of individual human action.
ABCT is an extension of this. It does not predict specific price levels or spending levels, or production levels or employment levels. Claiming that what took place during a specific period in time in history, like 1933, somehow refutes Austrian theory, shows that after all these years, you are still clueless about it.
But let’s be irrationally charitable and grant that 1933 to 1937 and 2009 to 2015 are inconsistent with what Austrian theory (does not) predict.
OK, then that would imply the other 2000 years out of the last 2015 years have been consistent with Austrian theory. That is over 99% accuracy.
Can you say cherry picking?
30. August 2015 at 13:23
Sumner wrote:
“I base it on Austrians who claim money was easy in the early 2000s because interest rates were low. Or that Austrian economists can somehow see foolish malinvestment occurring that markets are unable to see. That sort of thing.”
What do you mean that sort of thing?
You mean the sort of thing where you were somehow seeing foolish recession causing aggregate spending decreasing that the markets weren’t able to see until it was too late?
That the markets are so stupid when it comes to setting prices and spending on labor that there has to be some centralized non-market institution that prints unlimited quantities of money so that total spending does not fall?
That the markets are so stupid that they need you to convince them to enact NGDP futures betting where for hundreds of years prior the markets found no value in it?
Your fake kowtowing to the market is fake fake fake. You’re not pro market. You’re anti market as long as what is anti-market is what you prefer the anti-market activity to be.
News flash: This is the case even with self-professed socialists.
Market for me, socialism for thee.
30. August 2015 at 14:11
CNBC is now turning to RE to explain last Monday’s early 1000 point loss.
http://www.cnbc.com/2015/08/30/markets-plunge-may-actually-be-rational.html
31. August 2015 at 04:37
Levi, It’s been a while since I read any ABCT—why don’t you direct me to the best example, and let’s see what they can tell us about the macroeconomy.
31. August 2015 at 15:58
Scott,
I think Borio’s paper I linked above is quite good.
1. September 2015 at 12:28
Thanks Levi, It’s a useful reminder that the Austrianism that one observes on the internet is far from the best on offer. I certainly agree with those two papers on inflation targeting. But I still don’t find the overall framework to be a useful way of thinking about business cycles, even if they have some very good observations on particular issues.
I actually think I’d find Austrianism more appealing if I was a Keynesian, as that model has not done well in recent years. But I prefer the MM approach, that the recession made the financial crisis worse, to the Borio claim that the financial crisis made the recession worse.