Anrew Barrett review of one of our columnists Vitaliy Katsenelson's new book, Active Value Investing. Book Amazon subtitle: ‘Valuable data’ 4/5. 2007 Wiley Finance, 295 pages (of which 256 pages form the main body of the book)
________________________
Before I start this review, you should know that I didn't just buy a copy of this book, read and then review it (as I've done with all my other reviews). The author (who I'd not come across before) contacted me, asked if I would review his book and supplied me with a copy. As it was endorsed by Nassim Taleb and James Montier (both of whose work I admire), I thought it might be worth reading.
I've deliberately not read any other reviews or comments on the book and Katsenelson did not see this review before it was published. (When I worked as a sell-side analyst in the City (of London ) I absolutely deplored the practice of sending draft research notes to company management for them to check for 'errors' before publication.) On the other hand, saying good things about someone you've started to get to know is a lot easier than the opposite. I'll try to stay objective but you should be aware of my potential bias.
I was a bit sceptical of the book's title: surely value investing is value investing and the 'active' bit must be a gimmick? After reading the book I've come around, mainly won over by the extensive and very interesting statistics. Apart from the book's value in providing revision and reinforcement of the key ideas for value investing - which is always useful as we are generally exposed to plenty of bad influences that can screw up our thinking – it presents two particularly useful ideas. The first is that average stock market returns don't happen very often and the second is Katsenelson's multi-input PER (price/earnings ratio) model.
The first of these points is Katsenelson's main thesis: very long term (100 years+) average stock market returns (in the US) comprised protracted periods of above average returns (bull markets) followed by similarly long periods of below average returns (what Katsenelson calls 'range-bound' markets). I like the way the author puts it:
"…investors expecting the average returns observed over the past century are likely to be disappointed, as average happens a lot less frequently than we've been told."
This brings us to the most useful part of Katsenelson's book. His examination of data going back through several of these long-term cycles shows that economic growth, interest rates and inflation didn't explain the different returns in the periods of above average vs. below average returns. It was the starting valuation (PER) that mattered. The expansion or contraction of the market PER was responsible for virtually all of the difference in returns (with the exception of the Great Crash of 1929-1932) in the different market phases over the past century. I found this extremely interesting: I knew valuation mattered a lot, but I didn't know it was likely the only thing that matters (short of utter disaster).
Following on from this, Katsenelson attempts to show where we are in the cycles of above vs. below average returns (fortunately he understands that the most one can say at any point in time is that a certain outcome has a higher probability than other outcomes – I detest false certainty). I liked the way he approached this. Rather than using only, say, the one year historic PER, he presents the data in different ways: using the one year trailing PER and then also using the three, five and ten year average trailing PERs. This provides a useful sensitivity analysis to adjust for the current extremely high return on equity in the US and also allows us to draw our own conclusions if our opinion differs from that of the author (I think this is a great idea).
His analysis shows that we are very likely within a range-bound market that started in 2000, leading to two important practical investment considerations. The first is that dividends are critically important: they accounted for 90% of the average 5.9% nominal annual return during the range-bound markets Katsenelson has identified over the past century! The second is that being fully invested is much less important than in bull markets because, though the market fluctuates significantly in range-bound markets, the fluctuations cancel out. This advice chimes with that of, say, Mohnish Pabrai (who demonstrated with impeccable logic here: http://www.gurufocus.com/news.php?id=8955 and here: http://www.gurufocus.com/news.php?id=10005 that long-term buy and hold investing is extremely unlikely to generate returns much above 15% per annum).
Pabrai comes to the same rejection of the long-term buy and hold approach in The Dhandho Investor from the perspective of seeking the highest possible returns (without reference to the market). Katsenelson, however, believes that only substantial outperformance will produce satisfactory returns owing to the overall market's likely poor returns. Thus they both agree that an investor has to learn to sell (which many super-investors, including Marty Whitman and Joel Greenblatt consider very difficult or impossible to do well). Interestingly, Pabrai and Katsenelson both agree on the general principle: that you should have your exit plan in place before you invest.
This brings us on to Katsenelson's multi-input PER model, where he suggests using a simple PER model that adjusts an 'average' PER for such factors as growth, business quality, financial risk and dividend yield. I think this is a very good idea and is something I intend to try out. I had written some articles for a financial magazine in the UK a few years ago in which I'd suggested simply deducting the dividend yield from your required return, as it appears that dividends offer a 'free' source of return. (Counter-intuitively, high yield doesn't seem to reduce earnings growth according to a study by Cliff Asness and Rob Arnott that I also mentioned in my articles.) However, Katsenelson has taken the idea further and suggests a good (and most importantly, simple) way of comparing different companies and setting target sell prices.
I didn't like all of Katsenelson's book. For example, I found his effort to explain discounted cash flow analysis, using Tevye the milkman and his cow, somewhat confusing and I spotted a higher number of errors than normal (though I'm not sure if I was emailed the final version of the book).
The author's general conclusions about future US stock market returns have also already been presented by Warren Buffett in two articles published in Fortune magazine in 1999 and 2001 ("Warren Buffett on the stock market" by Carol Loomis, which can be found here: http://money.cnn.com/magazines/fortune/fortune_archive/1999/11/22/269071/index.htm and here: http://money.cnn.com/magazines/fortune/fortune_archive/2001/12/10/314691/index.htm).
Buffett and Katsenelson differ in their view of the importance of interest rates in affecting historical returns and Katsenelson (necessarily, as his is a book) presents considerably more detailed statistics. I'm also not sure that Buffett would believe that most investors would obtain any benefit from efforts to turn over their portfolios faster (the vast majority of investors have the opposite problem – as Katsenelson himself shows when he quotes a study showing the absurdly awful returns mutual fund investors actually achieved compared to the overall market in the 19 years to 2002).
So, where does that leave us? With a book containing some good ideas and excellent data and statistics but with a central conclusion (do more selling) that I suspect most investors will simply find too difficult to do well (notwithstanding Katsenelson's advice on how we might do so). Problems that originate from our psychological biases are very difficult to deal with satisfactorily: sometimes our efforts to improve (returns) can have the opposite effect.
________________________
Before I start this review, you should know that I didn't just buy a copy of this book, read and then review it (as I've done with all my other reviews). The author (who I'd not come across before) contacted me, asked if I would review his book and supplied me with a copy. As it was endorsed by Nassim Taleb and James Montier (both of whose work I admire), I thought it might be worth reading.
I've deliberately not read any other reviews or comments on the book and Katsenelson did not see this review before it was published. (When I worked as a sell-side analyst in the City (of London ) I absolutely deplored the practice of sending draft research notes to company management for them to check for 'errors' before publication.) On the other hand, saying good things about someone you've started to get to know is a lot easier than the opposite. I'll try to stay objective but you should be aware of my potential bias.
I was a bit sceptical of the book's title: surely value investing is value investing and the 'active' bit must be a gimmick? After reading the book I've come around, mainly won over by the extensive and very interesting statistics. Apart from the book's value in providing revision and reinforcement of the key ideas for value investing - which is always useful as we are generally exposed to plenty of bad influences that can screw up our thinking – it presents two particularly useful ideas. The first is that average stock market returns don't happen very often and the second is Katsenelson's multi-input PER (price/earnings ratio) model.
The first of these points is Katsenelson's main thesis: very long term (100 years+) average stock market returns (in the US) comprised protracted periods of above average returns (bull markets) followed by similarly long periods of below average returns (what Katsenelson calls 'range-bound' markets). I like the way the author puts it:
"…investors expecting the average returns observed over the past century are likely to be disappointed, as average happens a lot less frequently than we've been told."
This brings us to the most useful part of Katsenelson's book. His examination of data going back through several of these long-term cycles shows that economic growth, interest rates and inflation didn't explain the different returns in the periods of above average vs. below average returns. It was the starting valuation (PER) that mattered. The expansion or contraction of the market PER was responsible for virtually all of the difference in returns (with the exception of the Great Crash of 1929-1932) in the different market phases over the past century. I found this extremely interesting: I knew valuation mattered a lot, but I didn't know it was likely the only thing that matters (short of utter disaster).
Following on from this, Katsenelson attempts to show where we are in the cycles of above vs. below average returns (fortunately he understands that the most one can say at any point in time is that a certain outcome has a higher probability than other outcomes – I detest false certainty). I liked the way he approached this. Rather than using only, say, the one year historic PER, he presents the data in different ways: using the one year trailing PER and then also using the three, five and ten year average trailing PERs. This provides a useful sensitivity analysis to adjust for the current extremely high return on equity in the US and also allows us to draw our own conclusions if our opinion differs from that of the author (I think this is a great idea).
His analysis shows that we are very likely within a range-bound market that started in 2000, leading to two important practical investment considerations. The first is that dividends are critically important: they accounted for 90% of the average 5.9% nominal annual return during the range-bound markets Katsenelson has identified over the past century! The second is that being fully invested is much less important than in bull markets because, though the market fluctuates significantly in range-bound markets, the fluctuations cancel out. This advice chimes with that of, say, Mohnish Pabrai (who demonstrated with impeccable logic here: http://www.gurufocus.com/news.php?id=8955 and here: http://www.gurufocus.com/news.php?id=10005 that long-term buy and hold investing is extremely unlikely to generate returns much above 15% per annum).
Pabrai comes to the same rejection of the long-term buy and hold approach in The Dhandho Investor from the perspective of seeking the highest possible returns (without reference to the market). Katsenelson, however, believes that only substantial outperformance will produce satisfactory returns owing to the overall market's likely poor returns. Thus they both agree that an investor has to learn to sell (which many super-investors, including Marty Whitman and Joel Greenblatt consider very difficult or impossible to do well). Interestingly, Pabrai and Katsenelson both agree on the general principle: that you should have your exit plan in place before you invest.
This brings us on to Katsenelson's multi-input PER model, where he suggests using a simple PER model that adjusts an 'average' PER for such factors as growth, business quality, financial risk and dividend yield. I think this is a very good idea and is something I intend to try out. I had written some articles for a financial magazine in the UK a few years ago in which I'd suggested simply deducting the dividend yield from your required return, as it appears that dividends offer a 'free' source of return. (Counter-intuitively, high yield doesn't seem to reduce earnings growth according to a study by Cliff Asness and Rob Arnott that I also mentioned in my articles.) However, Katsenelson has taken the idea further and suggests a good (and most importantly, simple) way of comparing different companies and setting target sell prices.
I didn't like all of Katsenelson's book. For example, I found his effort to explain discounted cash flow analysis, using Tevye the milkman and his cow, somewhat confusing and I spotted a higher number of errors than normal (though I'm not sure if I was emailed the final version of the book).
The author's general conclusions about future US stock market returns have also already been presented by Warren Buffett in two articles published in Fortune magazine in 1999 and 2001 ("Warren Buffett on the stock market" by Carol Loomis, which can be found here: http://money.cnn.com/magazines/fortune/fortune_archive/1999/11/22/269071/index.htm and here: http://money.cnn.com/magazines/fortune/fortune_archive/2001/12/10/314691/index.htm).
Buffett and Katsenelson differ in their view of the importance of interest rates in affecting historical returns and Katsenelson (necessarily, as his is a book) presents considerably more detailed statistics. I'm also not sure that Buffett would believe that most investors would obtain any benefit from efforts to turn over their portfolios faster (the vast majority of investors have the opposite problem – as Katsenelson himself shows when he quotes a study showing the absurdly awful returns mutual fund investors actually achieved compared to the overall market in the 19 years to 2002).
So, where does that leave us? With a book containing some good ideas and excellent data and statistics but with a central conclusion (do more selling) that I suspect most investors will simply find too difficult to do well (notwithstanding Katsenelson's advice on how we might do so). Problems that originate from our psychological biases are very difficult to deal with satisfactorily: sometimes our efforts to improve (returns) can have the opposite effect.
