Is the bear dead? Maybe – at the very least the market had gotten insanely oversold and needed a bounce. In the bigger picture:
BULLISH: Copper prices have turned up noticeably over the past seven days. It is not a big move yet, but if it continues that is very bullish. Note that copper prices turned up a week before yesterday’s stock market bump.
http://www.kitcometals.com/charts/copper_historical_large.html
BEARISH: Economic news is still 100% bad (e.g. UTX news, WSJ p. B1). I do not count C’s announcement yesterday of operating profit in February as good economic news. They borrow from the fed at about 0% and lend at much greater than 0%. A monkey can make an operating profit in those conditions – they will still show a loss this quarter after marks and reserves. The fact that the market moved up so big on this silly news just shows how oversold the market was. Also bearish last week was increasing LIBOR and widening spreads in the debt markets.
BOTTOM LINE: It doesn’t feel over to me. It is hard for me to believe that such a ridiculous piece of “news” as Vikram Pandit’s remarks at Citi marked the end of the bear market. On the other hand, copper has been a very good indicator of market bottoms in the past, and if it keeps moving up then I will have to concede that the bear is probably dead. Also, if auto sales stabilize or turn up in March then that would also be very bullish (note they will report a huge drop vs. prior year’s March – the key is what happens to the seasonally adjusted annualized run-rate in March 09 vs. Feb. 09). Also watch for any other glimmers of economic good news.
Wednesday, March 11, 2009
Wednesday, March 4, 2009
Anatomy of the Bear
In his book, “Anatomy of the Bear,” Russell Napier studied the four greatest market bottoms of all time. He defined these by looking at the best returns over subsequent one to forty year time periods. In other words, he attempted identify the greatest buying opportunities of all time. Unfortunately, the lessons he found can not easily be distilled down into one short paragraph or three bullet points (perhaps some of the market’s current problems resulted from the expectation that everything can be). But in this case the lessons of history can be summarized in eleven bullet points which can be read in about ten minutes, so bear with me (no pun intended).
The four great market bottoms that Napier studied were August 1921, July 1932, June 1949 and August 1982. These market bottoms were the once (or maybe twice) in a lifetime buying opportunities for the market as a whole. Scientific study of these four bottoms revealed the following characteristics, which may be helpful indicators of the bottom of a major bear market (although note that other, less significant, market bottoms were not studied in this book, and therefore may have different traits). It certainly feels like we are heading for another one of these major bottoms. Interestingly, the data shows that some of the conventional wisdom that is being parroted by the pundits on CNBC and in the WSJ is not supported by history. Here is what Napier found:
1. The “q Ratio” (or “Tobin’s Q” for the Yale professor that came up with it) equals the total market value of a company (equity plus debt) divided by the replacement cost of all of its assets (net of liabilities). For the market as a whole, the q Ratio was below 0.3x at all four major market bottoms. This was the strongest correlation to all four exact market bottoms. Unfortunately, even though one can find published values for Tobin’s Q from various sources, its calculation requires that one know the replacement cost of a company’s net assets. This involves a fair amount of subjectivity and at best is only available months after the fact. Book value is often used as replacement value, but changing accounting practices for both depreciation and intangibles mean that this measure is different now than in the past, and in any event is never available in real time. MY CONCLUSION: At the end of the third quarter of 2008, the ratio was at 0.76 according to one source. While equity values have fallen since then (lowering Q), book values have also been written down significantly since that time (lifting Q). In December of 2008, Napier himself was quoted by Bloomberg as saying that by his calculations, a Q-ratio of 0.3x implies an S&P500 of about 400. Ouch!
2. The four great bear markets either took a very long time (more than a decade) or a very severe price decline (89% in the Dow from 1929 to 1932) to reach a long-term secular market bottom. In the three long bear markets (excluding 1929-32) equities became cheap slowly, taking an average of 14 years to move from peak to trough q ratios. Each of the four bear markets was a period of the market (gradually or quickly) becoming cheaper relative to corporate profits (particularly cyclically adjusted profits). In addition, real and nominal GDP expanded significantly during all three long bear markets, but corporate profit growth lagged growth in GDP. As a result, at the bottoms, cyclically adjusted P/E’s had declined to extremely low levels (4.7x to 11.7x trailing average five year earnings), and companies had room to both grow earnings (to catch up to GDP growth) and to expand P/E ratios. MY CONCLUSION: The current secular bear market began in the spring of 2000, so by this criteria, the current bear market (at 19 years old) may be nearing an end. Furthermore, by my own personal calculation, at 700 the S&P500 is trading at about 10.8x current normalized earnings (based on along term regression analysis), vs. a “normal” level of perhaps 15x. However, at its high in fall 2007 the ratio was 25x, or 1.67 above the 15x norm. If the market overreacts by a similar amount the other way, that implies a multiple of 9x (15 divided by 1.67), which based on 2009 normalized earnings implies an S&P500 of about 600. Ouch again!
3. In all four cases, the market did NOT bottom with “capitulation” of the bulls (where there is a final dramatic decline on high volume) as conventional wisdom today dictates. Instead, it was more like a marathon, with the weaker bulls gradually dropping out through attrition by the side of the road over a period of months or years. The bottoms were all preceded by a period of declines on low volumes and rises on higher volumes (although not necessarily that high in absolute terms). The final end of the bear in each case was a slump in equity prices on low volumes. After that, the new long-term up trend was confirmed by rising volumes at new higher price levels after the initial market rebound. MY CONCLUSION: How interesting that everyone on CNBC is looking for capitulation – a huge downdraft on massive volume. If this market is like the other four, that is NOT how this bear will end. The volume data for the NYSE has not really shown a sustained decline at this point. If volume trails off and then picks up again, this would be bullish. (Monthly S&P500 volume can be found at http://www.Economagic.com/chartg/sp/sp05.gif)
4. A material disturbance to the general price level was the catalyst to reduce equity prices to extremely cheap levels (high inflation and then deflation leading up to 1921, 1949 and 1982; just extreme deflation up to 1932). The deflation was particularly evident in declining commodity prices. This caused great uncertainty, reducing valuations. In all four cases, the end of the decline in commodity prices, particularly copper, marked the bottom of the great bear market. MY CONCLUSION: That first sentence certainly replayed itself leading up to this bust (think oil boom and bust, among others). Copper traded around $1/lb in 2004 and early 2005, and then in late 2005 and 2006 moved above $3. For most of 2006 through mid-2008 it traded between $3 and $4. In the last few months of 2008 it fell rapidly to around $1.50, where it has stayed relatively level ever since. Since Copper appears to have bottomed, this could signal this bear is about done. A new downturn would be bearish; an upturn would be even more bullish. A good source for copper prices is http://www.kitcometals.com/charts/copper_historical_large.html
5. All four bottoms occurred during a recession. Corporate earnings continued to decline for four to seven months after the market bottomed. In hindsight, in each case the overall economy (GDP) and the stock market bottomed at roughly the same time, but this was not much help to investors (given the lag and then subsequent revisions in the reporting of economic data). However, reports of auto sales starting to recover preceded each bear market bottom. MY CONCLUSION: Earnings and GDP known only in hindsight are not much help, but watch auto sales for an upturn, especially after the recent brutal downturn. February marked the lowest annualized auto sales rate (9.12 million vehicles) in the US since December 1981. I don’t have a good source for a nice chart on this, but the latest data can generally be found at http://www.aiada.org/newsroom/ or http://www.autoblog.com/category/by-the-numbers/.
6. At each bottom there was an increasing supply of good economic news and a significant number of bulls “banging the drum” for equities, but the market was indifferent to all of it. Conventional wisdom that says that the news at the bottom is universally bad is wrong. At the bottoms it was NOT true that all the news was bad, rather it was true that the market had continued to ignore an increasing amount of good news. MY CONCLUSION: There are a few on CNBC who think the time to buy is now, but very few. And there is pretty much no good economic news yet.
7. Also at the bottoms there were many commentators suggesting that the government’s worsening fiscal position (budget deficits, etc.) would prevent an increase in stocks (they were obviously wrong). MY CONCLUSION: That is sounding more familiar everyday.
8. Large numbers of individuals were shorting stocks at each bottom. Short positions increased to higher levels in the early stages of each bear market’s ultimate recovery, as shorts assumed it was a new opportunity to short at higher levels before the next decline (they were obviously wrong too). MY CONCLUSION: This is not that much help, since it is a bit vague and since good short interest data for the market as a whole is difficult to find, and is further clouded by the advent of short and ultra-short ETF’s. But it is interesting that this goes counter to the notion that short-covering triggers a recovery, since initially in the recoveries short interest increased.
9. Dow Theory worked as a good buy signal at each bottom, although Dow Theory is itself subject to a fair amount of subjectivity and interpretation. MY CONCLUSION: As Napier himself says this is somewhat subjective, but a move up in the Dow Industrials and Transports, together and on high volume, would be bullish.
10. Ignoring the 1929-1932 bear, the Federal Reserve’s first reduction in interest rates preceded the bottom by three to 11 months, during which time the market declined a further 10% - 20%. However, in 1929 this indicator failed miserably as the Fed cut interest rates very early in the bear market. MY CONCLUSION: This is probably not very reliable this time either, since (as in 1929) the Fed again cut hard and early in this recession.
11. A sell off in government bonds accompanied at least part of each bear market in equities. The bear market in bonds ended seven to 14 months before the end of the equity bear market, and in that time the equity market declined 6% to 42%. MY CONCLUSION: So far in this bear market, US Treasury Bonds have not been in a bear market, as ten-year notes traded to absurdly high levels (low yields) in late 2008. They have given some of it back so far in 2009, but yields are still quite low (prices are still higher than they were up until the fall of 2008). There are two possible interpretations of this. One is that Treasury Notes still need to suffer a bear market before the equity bear market ends. I do not think this is the case, since for that to happen, and then for the equity bear market to go on another seven to 14 months beyond that, seems unreasonable. A more likely explanation is that Treasury Notes have been a safe haven that have behaved more like gold in a crisis, and so you can’t really look for a bear market there. Certainly there has been a bear market in high yield corporates, as spreads to Treasuries widened significantly. For example, PIMCO’s high yield ETF (ticker PHK) had asset values begin to decline in mid-2007 until they bottomed in late 2008, down about 70% in 18 months (http://www.etfconnect.com/select/fundpages/gen.asp?MFID=108698). They have since stabilized, although not rebounded.
For what it’s worth, the author ended the book with his assessment of where we stood at the time in the fall of 2005. At that time, the market had not surpassed its 2000 highs, and his conclusion was that the rebound of 2003 – 2005 was just a short term (cyclical) bull market in a long term (secular) bear market. He pointed out that historically overvalued levels like we saw in 2000 are always followed by long grinding bear markets (or as in 1929 – 1932, by a not so long but really extremely severe bear market), which don’t finally end until equities become extremely undervalued.
Also, for what it's worth, I will add one more metric of my own. In mid-1982, the ratio of the S&P500 index to quarterly GDP (in trillions, seasonally adjusted) approached 32x. 32x fourth quarter 2008 GDP is approximately 450 for the S&P500. Double ouch! However in 1949 the level did not get nearly as low – in fact it bottomed around 50x, and 50x last quarters GDP works out to 710 for the S&P500, right about the current level. (GDP data can be found at http://www.bea.gov/national/nipaweb/TableView.asp?SelectedTable=5&FirstYear=2007&LastYear=2008&Freq=Qtr).
CONCLUSIONS:
The data seems to indicate the current bear market should bottom somewhere in the 400 to 700 range for the S&P500.
Signs that the bear is dead:
A lethargic slump in equity prices on low volumes, followed by rising volumes at new higher price levels after the initial market rebound.
An upturn in Copper prices.
An upturn in Auto Sales.
An increasing supply of good economic news (initially ignored by the market).
Signs that the bear has a while to live:
Falling Copper prices
Flat to down auto sales.
All the economic news is bad.
The four great market bottoms that Napier studied were August 1921, July 1932, June 1949 and August 1982. These market bottoms were the once (or maybe twice) in a lifetime buying opportunities for the market as a whole. Scientific study of these four bottoms revealed the following characteristics, which may be helpful indicators of the bottom of a major bear market (although note that other, less significant, market bottoms were not studied in this book, and therefore may have different traits). It certainly feels like we are heading for another one of these major bottoms. Interestingly, the data shows that some of the conventional wisdom that is being parroted by the pundits on CNBC and in the WSJ is not supported by history. Here is what Napier found:
1. The “q Ratio” (or “Tobin’s Q” for the Yale professor that came up with it) equals the total market value of a company (equity plus debt) divided by the replacement cost of all of its assets (net of liabilities). For the market as a whole, the q Ratio was below 0.3x at all four major market bottoms. This was the strongest correlation to all four exact market bottoms. Unfortunately, even though one can find published values for Tobin’s Q from various sources, its calculation requires that one know the replacement cost of a company’s net assets. This involves a fair amount of subjectivity and at best is only available months after the fact. Book value is often used as replacement value, but changing accounting practices for both depreciation and intangibles mean that this measure is different now than in the past, and in any event is never available in real time. MY CONCLUSION: At the end of the third quarter of 2008, the ratio was at 0.76 according to one source. While equity values have fallen since then (lowering Q), book values have also been written down significantly since that time (lifting Q). In December of 2008, Napier himself was quoted by Bloomberg as saying that by his calculations, a Q-ratio of 0.3x implies an S&P500 of about 400. Ouch!
2. The four great bear markets either took a very long time (more than a decade) or a very severe price decline (89% in the Dow from 1929 to 1932) to reach a long-term secular market bottom. In the three long bear markets (excluding 1929-32) equities became cheap slowly, taking an average of 14 years to move from peak to trough q ratios. Each of the four bear markets was a period of the market (gradually or quickly) becoming cheaper relative to corporate profits (particularly cyclically adjusted profits). In addition, real and nominal GDP expanded significantly during all three long bear markets, but corporate profit growth lagged growth in GDP. As a result, at the bottoms, cyclically adjusted P/E’s had declined to extremely low levels (4.7x to 11.7x trailing average five year earnings), and companies had room to both grow earnings (to catch up to GDP growth) and to expand P/E ratios. MY CONCLUSION: The current secular bear market began in the spring of 2000, so by this criteria, the current bear market (at 19 years old) may be nearing an end. Furthermore, by my own personal calculation, at 700 the S&P500 is trading at about 10.8x current normalized earnings (based on along term regression analysis), vs. a “normal” level of perhaps 15x. However, at its high in fall 2007 the ratio was 25x, or 1.67 above the 15x norm. If the market overreacts by a similar amount the other way, that implies a multiple of 9x (15 divided by 1.67), which based on 2009 normalized earnings implies an S&P500 of about 600. Ouch again!
3. In all four cases, the market did NOT bottom with “capitulation” of the bulls (where there is a final dramatic decline on high volume) as conventional wisdom today dictates. Instead, it was more like a marathon, with the weaker bulls gradually dropping out through attrition by the side of the road over a period of months or years. The bottoms were all preceded by a period of declines on low volumes and rises on higher volumes (although not necessarily that high in absolute terms). The final end of the bear in each case was a slump in equity prices on low volumes. After that, the new long-term up trend was confirmed by rising volumes at new higher price levels after the initial market rebound. MY CONCLUSION: How interesting that everyone on CNBC is looking for capitulation – a huge downdraft on massive volume. If this market is like the other four, that is NOT how this bear will end. The volume data for the NYSE has not really shown a sustained decline at this point. If volume trails off and then picks up again, this would be bullish. (Monthly S&P500 volume can be found at http://www.Economagic.com/chartg/sp/sp05.gif)
4. A material disturbance to the general price level was the catalyst to reduce equity prices to extremely cheap levels (high inflation and then deflation leading up to 1921, 1949 and 1982; just extreme deflation up to 1932). The deflation was particularly evident in declining commodity prices. This caused great uncertainty, reducing valuations. In all four cases, the end of the decline in commodity prices, particularly copper, marked the bottom of the great bear market. MY CONCLUSION: That first sentence certainly replayed itself leading up to this bust (think oil boom and bust, among others). Copper traded around $1/lb in 2004 and early 2005, and then in late 2005 and 2006 moved above $3. For most of 2006 through mid-2008 it traded between $3 and $4. In the last few months of 2008 it fell rapidly to around $1.50, where it has stayed relatively level ever since. Since Copper appears to have bottomed, this could signal this bear is about done. A new downturn would be bearish; an upturn would be even more bullish. A good source for copper prices is http://www.kitcometals.com/charts/copper_historical_large.html
5. All four bottoms occurred during a recession. Corporate earnings continued to decline for four to seven months after the market bottomed. In hindsight, in each case the overall economy (GDP) and the stock market bottomed at roughly the same time, but this was not much help to investors (given the lag and then subsequent revisions in the reporting of economic data). However, reports of auto sales starting to recover preceded each bear market bottom. MY CONCLUSION: Earnings and GDP known only in hindsight are not much help, but watch auto sales for an upturn, especially after the recent brutal downturn. February marked the lowest annualized auto sales rate (9.12 million vehicles) in the US since December 1981. I don’t have a good source for a nice chart on this, but the latest data can generally be found at http://www.aiada.org/newsroom/ or http://www.autoblog.com/category/by-the-numbers/.
6. At each bottom there was an increasing supply of good economic news and a significant number of bulls “banging the drum” for equities, but the market was indifferent to all of it. Conventional wisdom that says that the news at the bottom is universally bad is wrong. At the bottoms it was NOT true that all the news was bad, rather it was true that the market had continued to ignore an increasing amount of good news. MY CONCLUSION: There are a few on CNBC who think the time to buy is now, but very few. And there is pretty much no good economic news yet.
7. Also at the bottoms there were many commentators suggesting that the government’s worsening fiscal position (budget deficits, etc.) would prevent an increase in stocks (they were obviously wrong). MY CONCLUSION: That is sounding more familiar everyday.
8. Large numbers of individuals were shorting stocks at each bottom. Short positions increased to higher levels in the early stages of each bear market’s ultimate recovery, as shorts assumed it was a new opportunity to short at higher levels before the next decline (they were obviously wrong too). MY CONCLUSION: This is not that much help, since it is a bit vague and since good short interest data for the market as a whole is difficult to find, and is further clouded by the advent of short and ultra-short ETF’s. But it is interesting that this goes counter to the notion that short-covering triggers a recovery, since initially in the recoveries short interest increased.
9. Dow Theory worked as a good buy signal at each bottom, although Dow Theory is itself subject to a fair amount of subjectivity and interpretation. MY CONCLUSION: As Napier himself says this is somewhat subjective, but a move up in the Dow Industrials and Transports, together and on high volume, would be bullish.
10. Ignoring the 1929-1932 bear, the Federal Reserve’s first reduction in interest rates preceded the bottom by three to 11 months, during which time the market declined a further 10% - 20%. However, in 1929 this indicator failed miserably as the Fed cut interest rates very early in the bear market. MY CONCLUSION: This is probably not very reliable this time either, since (as in 1929) the Fed again cut hard and early in this recession.
11. A sell off in government bonds accompanied at least part of each bear market in equities. The bear market in bonds ended seven to 14 months before the end of the equity bear market, and in that time the equity market declined 6% to 42%. MY CONCLUSION: So far in this bear market, US Treasury Bonds have not been in a bear market, as ten-year notes traded to absurdly high levels (low yields) in late 2008. They have given some of it back so far in 2009, but yields are still quite low (prices are still higher than they were up until the fall of 2008). There are two possible interpretations of this. One is that Treasury Notes still need to suffer a bear market before the equity bear market ends. I do not think this is the case, since for that to happen, and then for the equity bear market to go on another seven to 14 months beyond that, seems unreasonable. A more likely explanation is that Treasury Notes have been a safe haven that have behaved more like gold in a crisis, and so you can’t really look for a bear market there. Certainly there has been a bear market in high yield corporates, as spreads to Treasuries widened significantly. For example, PIMCO’s high yield ETF (ticker PHK) had asset values begin to decline in mid-2007 until they bottomed in late 2008, down about 70% in 18 months (http://www.etfconnect.com/select/fundpages/gen.asp?MFID=108698). They have since stabilized, although not rebounded.
For what it’s worth, the author ended the book with his assessment of where we stood at the time in the fall of 2005. At that time, the market had not surpassed its 2000 highs, and his conclusion was that the rebound of 2003 – 2005 was just a short term (cyclical) bull market in a long term (secular) bear market. He pointed out that historically overvalued levels like we saw in 2000 are always followed by long grinding bear markets (or as in 1929 – 1932, by a not so long but really extremely severe bear market), which don’t finally end until equities become extremely undervalued.
Also, for what it's worth, I will add one more metric of my own. In mid-1982, the ratio of the S&P500 index to quarterly GDP (in trillions, seasonally adjusted) approached 32x. 32x fourth quarter 2008 GDP is approximately 450 for the S&P500. Double ouch! However in 1949 the level did not get nearly as low – in fact it bottomed around 50x, and 50x last quarters GDP works out to 710 for the S&P500, right about the current level. (GDP data can be found at http://www.bea.gov/national/nipaweb/TableView.asp?SelectedTable=5&FirstYear=2007&LastYear=2008&Freq=Qtr).
CONCLUSIONS:
The data seems to indicate the current bear market should bottom somewhere in the 400 to 700 range for the S&P500.
Signs that the bear is dead:
A lethargic slump in equity prices on low volumes, followed by rising volumes at new higher price levels after the initial market rebound.
An upturn in Copper prices.
An upturn in Auto Sales.
An increasing supply of good economic news (initially ignored by the market).
Signs that the bear has a while to live:
Falling Copper prices
Flat to down auto sales.
All the economic news is bad.
Subscribe to:
Posts (Atom)
Posts
