Central Banks Can't Target Asset Prices; Reversion To The Mean May Not Work
I ran across a blog entry at Aleph referring to a Bloomberg story suggesting that some believe asset prices should be targetted by central banks. Mark Thoma also tackles this topic in a SeekingAlpha post. IANAE (I'm not an economist) but here is my layperson perspective on this.
The Bloomberg story is a bit narrow and concerned with asset bubbles due to excessive credit (basically the current problem.) But if we generalize it and consider asset bubbles in general, I would say it's impossible for central banks to target asset bubbles. I quote a WSJ blog entry last year dealing with this topic so you may want to check that out. It's impossible to know what is in a bubble ahead of time. For instance, consider some of the recent scenarios.
People like me entertained the possibility that commodities were in a bubble, yet we have many others who are confident that commodities are not in a bubble. So which is it? The answer, of course, is that we don't know right now. But in 10 years we will. If commodities collapse and don't rise back up then the recent environment was clearly a bubble. Until then, it's hard to say.
Similarly, consider the technology bubble in the last 90's. Right now, everyone accepts the view that technology--more accurately TMT or technology, media, and telecommunications--was in a big bubble by 2000. But could anyone have known in, say, 1996 or 1997 that there was a bubble developing? The answer would have been negative. Consider the following table from Bespoke Investment Group showing the sector share of the S&P 500 (source: SeekingAlpha):
I will note that market cap does not necessarily reflect the actual share of the economy but I'm just using it as a crude approximation. What you would have seen in 1997 is a situation where technology is around 12% of the S&P 500. It was much lower in the past so you actually may have felt it was in a bubble. You would have seen the whole industry booming, with shares of companies like Microsoft seemingly rising every month, in a bubble-like fashion.
If governments had a policy of targetting bubbles, they may have crushed the technology sector in 1996 or 1997. Yet that would have been a disastrous policy for the country. What you were observing was not a bubble in formation but the emergence of a new set of industries. Technology was clearly in a bubble by 2000 but I would argue it was not in 1996 or 1997. In fact, even after the tech bust, the technology share of S&P 500 never went back anywhere near the 12% in 1997 (it has stayed above 15% all throughout the 2000's and likely may for decades.)
I hope you can see why it's hard to tell what is a bubble in advance or while you are in the middle of it. Similar to the technology scenario in the 1990's, it's very difficult to say if commodities are in a bubble. Remember that this is coming from some guy who was bearish on commodities for the last few years and even put his money on the line (too bad it didn't create much wealth : ). As bearish as I have been, it's difficult to say if this is a long-term bubble. Emerging countries have been developing rapidly so we may be witnessing the creation of new industries. Or we may not; all this could just be hype. We just don't know.
I hope that sort of illustrates why bubbles are hard to identify. The commodities example is a very good one. No one really knows if oil at $140 is a bubble or if $100 is a bubble. Given the reality of the situation, I don't believe any government body can target assets. If they do so, they will kill off every emerging industry that can actually create wealth for the country.
Reversion To The Mean Can Be Misleading
The technology example above, along with the data from Bespoke Investment Group, clearly illustrates why reversion to the mean can be misleading. If you were looking at technology in the mid-90's, you clearly would have thought it would decline. Technology was a smaller part of the economy and the stock market before the 90's. You may have thought it would revert to the mean and shrink. Needless to say, you would have been completely wrong. Not only did technology expand, it's one of the largest sectors of the stock market and the US economy.
What is more likely to revert to the mean are valuations, not the share of the market. Technology may be 15% of the market cap now versus around 6% in 1990, but it's difficult to argue that it will decline back to 6%. Instead, valuation metrics such as P/E, P/BV, and so forth, are more likely to revert.
(One other item one should keep in mind is that the definition of sectors and industries changes over time. Radio would have been high-tech back in the 1920's but that's not the case now. Looking forward, businesses related to alternative energy may either end up being slotted into technology, or industrials, or utilities (the result can make a huge impact if there is a boom in this area.))
The Bloomberg story is a bit narrow and concerned with asset bubbles due to excessive credit (basically the current problem.) But if we generalize it and consider asset bubbles in general, I would say it's impossible for central banks to target asset bubbles. I quote a WSJ blog entry last year dealing with this topic so you may want to check that out. It's impossible to know what is in a bubble ahead of time. For instance, consider some of the recent scenarios.
People like me entertained the possibility that commodities were in a bubble, yet we have many others who are confident that commodities are not in a bubble. So which is it? The answer, of course, is that we don't know right now. But in 10 years we will. If commodities collapse and don't rise back up then the recent environment was clearly a bubble. Until then, it's hard to say.
Similarly, consider the technology bubble in the last 90's. Right now, everyone accepts the view that technology--more accurately TMT or technology, media, and telecommunications--was in a big bubble by 2000. But could anyone have known in, say, 1996 or 1997 that there was a bubble developing? The answer would have been negative. Consider the following table from Bespoke Investment Group showing the sector share of the S&P 500 (source: SeekingAlpha):
(source: Bespoke Investment Group)
I will note that market cap does not necessarily reflect the actual share of the economy but I'm just using it as a crude approximation. What you would have seen in 1997 is a situation where technology is around 12% of the S&P 500. It was much lower in the past so you actually may have felt it was in a bubble. You would have seen the whole industry booming, with shares of companies like Microsoft seemingly rising every month, in a bubble-like fashion.
If governments had a policy of targetting bubbles, they may have crushed the technology sector in 1996 or 1997. Yet that would have been a disastrous policy for the country. What you were observing was not a bubble in formation but the emergence of a new set of industries. Technology was clearly in a bubble by 2000 but I would argue it was not in 1996 or 1997. In fact, even after the tech bust, the technology share of S&P 500 never went back anywhere near the 12% in 1997 (it has stayed above 15% all throughout the 2000's and likely may for decades.)
I hope you can see why it's hard to tell what is a bubble in advance or while you are in the middle of it. Similar to the technology scenario in the 1990's, it's very difficult to say if commodities are in a bubble. Remember that this is coming from some guy who was bearish on commodities for the last few years and even put his money on the line (too bad it didn't create much wealth : ). As bearish as I have been, it's difficult to say if this is a long-term bubble. Emerging countries have been developing rapidly so we may be witnessing the creation of new industries. Or we may not; all this could just be hype. We just don't know.
I hope that sort of illustrates why bubbles are hard to identify. The commodities example is a very good one. No one really knows if oil at $140 is a bubble or if $100 is a bubble. Given the reality of the situation, I don't believe any government body can target assets. If they do so, they will kill off every emerging industry that can actually create wealth for the country.
Reversion To The Mean Can Be Misleading
The technology example above, along with the data from Bespoke Investment Group, clearly illustrates why reversion to the mean can be misleading. If you were looking at technology in the mid-90's, you clearly would have thought it would decline. Technology was a smaller part of the economy and the stock market before the 90's. You may have thought it would revert to the mean and shrink. Needless to say, you would have been completely wrong. Not only did technology expand, it's one of the largest sectors of the stock market and the US economy.
What is more likely to revert to the mean are valuations, not the share of the market. Technology may be 15% of the market cap now versus around 6% in 1990, but it's difficult to argue that it will decline back to 6%. Instead, valuation metrics such as P/E, P/BV, and so forth, are more likely to revert.
(One other item one should keep in mind is that the definition of sectors and industries changes over time. Radio would have been high-tech back in the 1920's but that's not the case now. Looking forward, businesses related to alternative energy may either end up being slotted into technology, or industrials, or utilities (the result can make a huge impact if there is a boom in this area.))
Comments
Post a Comment