My latest paper for GMO: Is Austerity the Road to Ruin can be found here:
http://www.gmo.com/
or
http://www.scribd.com/doc/34895648/GMO-Montier-26Jul
Very interesting response to this paper, I have rarely been called a moron more often.
Others have accused me of talking my book, presumably wanting QEII to drive the markets higher, frankly given the stocks we own the last thing we want is yet more speculative demand for junky stocks!
Tuesday, 27 July 2010
Thursday, 22 July 2010
On the price of insurance and the bull market in tail risk
From the people who bought you such wonderful ideas as CDOs, now comes the bull market in tail risk products. Deutsche are launching a long equity volatilty index, Citi has come up with a crisis index (mixing equity and bond vols, swap spreads and structured credit spreads). Bloomberg reports that PIMCO is planning a fund that will protect investors in the event of a decline greater than 15%. The CBOE is planning a new index based on the skew in the S&P500.
In the past I've talked about the need for cheap insurance, and the benefits that this can bring to a portfolio in terms of robustness. However, the key word is that the insurance must be cheap (or at very worst fair value). Buying expensive insurance is a waste of time. I used to live in Tokyo and was constantly amazed that the day after an earth tremor the cost of earthquake insurance would soar, as would the demand! You should really only want insurance when it is cheap, as this is the time when the no one else wants it, and (perversely) the events are most likely. Buying expensive insurance is just like buying any other overpriced asset...a path to the permanent impairment of captial. Rather than wasting money on expensive insurance, holding a larger cash balance makes sense. It preserves the dry powder for times when you want to deploy capital, and limits the downside.
So buy insurance when it's cheap. When it isn't and you are worried about the downside, hold cash. As Buffet said holding cash is painful, but not as painful as doing something stupid!
In the past I've talked about the need for cheap insurance, and the benefits that this can bring to a portfolio in terms of robustness. However, the key word is that the insurance must be cheap (or at very worst fair value). Buying expensive insurance is a waste of time. I used to live in Tokyo and was constantly amazed that the day after an earth tremor the cost of earthquake insurance would soar, as would the demand! You should really only want insurance when it is cheap, as this is the time when the no one else wants it, and (perversely) the events are most likely. Buying expensive insurance is just like buying any other overpriced asset...a path to the permanent impairment of captial. Rather than wasting money on expensive insurance, holding a larger cash balance makes sense. It preserves the dry powder for times when you want to deploy capital, and limits the downside.
So buy insurance when it's cheap. When it isn't and you are worried about the downside, hold cash. As Buffet said holding cash is painful, but not as painful as doing something stupid!
Wednesday, 7 July 2010
Barbie does economics
Barbie does economics!
The sheer hubris of many in the economics profession never ceases to amaze me. Take for instance a recent paper by Kartik Athreya of the Federal Reserve Bank of Richmond[1] entitled “Economics is Hard. Don’t let Bloggers Tell You Otherwise”. In a move that is eerily reminiscent of the controversial talking Barbie of the early 1990s who fatefully uttered “Math class is tough”[2], Athreya’s short paper essentially lays out a quite staggering claim :- that economics should be left to those with a PhD in the subject!
Athreya describes himself as “a worker bee chipping away with known tools”. He goes on to say “writers who have not taken a year of PhD coursework in a decent economics department…cannot meaningfully advance the discussion on economic policy”[3]. You’ve got to love the ‘decent’ in that sentence – it reeks of intellectual snobbishness of the highest order.
In fact, Athreya’s ire isn’t limited to what he sees as uninformed debate, he seems to object to anyone who attempts to make the policy issues of the day clear even if they have a PhD. He pejoratively describes both Paul Krugman and Brad DeLong as “Patron saints of the Macroeconomic Policy is Easy” movement.
He argues that we won’t expect particularly informed discussion on the causes, consequences and treatments for cancer from non-Oncology specialists, so why we would we expect non-specialists to offer any useful debate on economics.
However, the analogy is false. Modern medicine is based on scientific principles and follows an evidence based approach. Even then some estimate that the majority of published findings in medical journals are false[4]!
Economics starts from a far worse place. It isn’t a science, and often seems more interested in twisting the facts to fit a theory rather than the other way around. In fact, as Nassim Taleb has pointed out, economics is more akin to medieval medicine than its current practice, “Medicine used to kill more patients than it saved – just as financial economics endangers the system by creating, not reducing, risk.[5]”
The idea that what we need is more ‘worker bees’ gaining their PhD’s from conducting ‘angels on a pin head’ like work based on minor alterations to previous research makes me want to cry. Where were the warnings from the orthodox economics establishment ahead of the global financial crisis? Oh, that’s right there weren’t any.
Indeed many of those who warned of the problems ahead did so because they weren’t constrained by the kind of training that an economics PhD suffers. I did my own training in economics a long time ago now, it included a fair amount of equation bending but I was incredibly fortunate that it included generalist topics such as Marxian and post-Keynesian economics, subjects that are oddly absent from the vast majority of syllabi.
In many ways economics as it exists today is largely a victim of learned helplessness - a phenomenon was first documented by Martin Seligman in the 1960s. He was working with dogs (dog lovers look away now) and studying conditioning when he came across something interesting. Seligman was subjecting pairs of dogs to nondamaging but painful electric shocks. However, in each pair of dogs one animal could put an end to the shock by simply pressing the side panels of its container with its head. The other dog was unable to turn off the shock. The electricity was synchronized, starting at the same point for both dogs, and obviously ending when the dog with the control turned off the power.
This gave the each of the pairs of dogs a very different experience. One experienced the pain as controllable, while the other did not. The dogs which had no control soon began to cower and whine (signs of doggy depression) even after the sessions had stopped. The dogs which could control the shocks showed no signs of this behaviour.
In the second phase of the experiments dogs were placed in box with a low wall separating the container into two. One side (the side on which the dog started) was electrified. To avoid the pain the dog simply had to jump the low wall. The dogs which had controlled the shocks in the first round quickly learned to jump the wall. However, the around two-thirds of the dogs who had no control in the first round, simply laid down and suffered the pain, they had learned to become helpless.
Modern day economics is much like these poor animals. Many economists have learnt to become helpless. They would rather lay down and whimper and whine about how unfair the world is, and mutter that everything would be alright if only people behaved like their models, than seek to look outside the narrow confines of their obsession with rationality and mathematics to see if others might just have some useful insight.
The age of the specialist (people who learn more and more about less and less, until they know absolutely everything about nothing) has proved to have some fundamental flaws. Three cheers for the generalists!
[1] http://www.scribd.com/doc/33655771/Economics-is-Hard
[2] For more on the weird and wonderful versions of Barbie that have graced the shelves over the years see http://shine.yahoo.com/channel/parenting/11-bad-barbie-ideas-1312923/?pg=8
[3] Atherya does have the sense to point out “Taken literally, I am almost certainly wrong.”
[4] John Ioannidis (2005) Why Most Published Research Findings Are False. PLoS Med 2(8): e124. doi:10.1371/journal.pmed.0020124
[5] Taleb (2007) The pseudo science hurting markets, Financial Times, 23 October 2007
The sheer hubris of many in the economics profession never ceases to amaze me. Take for instance a recent paper by Kartik Athreya of the Federal Reserve Bank of Richmond[1] entitled “Economics is Hard. Don’t let Bloggers Tell You Otherwise”. In a move that is eerily reminiscent of the controversial talking Barbie of the early 1990s who fatefully uttered “Math class is tough”[2], Athreya’s short paper essentially lays out a quite staggering claim :- that economics should be left to those with a PhD in the subject!
Athreya describes himself as “a worker bee chipping away with known tools”. He goes on to say “writers who have not taken a year of PhD coursework in a decent economics department…cannot meaningfully advance the discussion on economic policy”[3]. You’ve got to love the ‘decent’ in that sentence – it reeks of intellectual snobbishness of the highest order.
In fact, Athreya’s ire isn’t limited to what he sees as uninformed debate, he seems to object to anyone who attempts to make the policy issues of the day clear even if they have a PhD. He pejoratively describes both Paul Krugman and Brad DeLong as “Patron saints of the Macroeconomic Policy is Easy” movement.
He argues that we won’t expect particularly informed discussion on the causes, consequences and treatments for cancer from non-Oncology specialists, so why we would we expect non-specialists to offer any useful debate on economics.
However, the analogy is false. Modern medicine is based on scientific principles and follows an evidence based approach. Even then some estimate that the majority of published findings in medical journals are false[4]!
Economics starts from a far worse place. It isn’t a science, and often seems more interested in twisting the facts to fit a theory rather than the other way around. In fact, as Nassim Taleb has pointed out, economics is more akin to medieval medicine than its current practice, “Medicine used to kill more patients than it saved – just as financial economics endangers the system by creating, not reducing, risk.[5]”
The idea that what we need is more ‘worker bees’ gaining their PhD’s from conducting ‘angels on a pin head’ like work based on minor alterations to previous research makes me want to cry. Where were the warnings from the orthodox economics establishment ahead of the global financial crisis? Oh, that’s right there weren’t any.
Indeed many of those who warned of the problems ahead did so because they weren’t constrained by the kind of training that an economics PhD suffers. I did my own training in economics a long time ago now, it included a fair amount of equation bending but I was incredibly fortunate that it included generalist topics such as Marxian and post-Keynesian economics, subjects that are oddly absent from the vast majority of syllabi.
In many ways economics as it exists today is largely a victim of learned helplessness - a phenomenon was first documented by Martin Seligman in the 1960s. He was working with dogs (dog lovers look away now) and studying conditioning when he came across something interesting. Seligman was subjecting pairs of dogs to nondamaging but painful electric shocks. However, in each pair of dogs one animal could put an end to the shock by simply pressing the side panels of its container with its head. The other dog was unable to turn off the shock. The electricity was synchronized, starting at the same point for both dogs, and obviously ending when the dog with the control turned off the power.
This gave the each of the pairs of dogs a very different experience. One experienced the pain as controllable, while the other did not. The dogs which had no control soon began to cower and whine (signs of doggy depression) even after the sessions had stopped. The dogs which could control the shocks showed no signs of this behaviour.
In the second phase of the experiments dogs were placed in box with a low wall separating the container into two. One side (the side on which the dog started) was electrified. To avoid the pain the dog simply had to jump the low wall. The dogs which had controlled the shocks in the first round quickly learned to jump the wall. However, the around two-thirds of the dogs who had no control in the first round, simply laid down and suffered the pain, they had learned to become helpless.
Modern day economics is much like these poor animals. Many economists have learnt to become helpless. They would rather lay down and whimper and whine about how unfair the world is, and mutter that everything would be alright if only people behaved like their models, than seek to look outside the narrow confines of their obsession with rationality and mathematics to see if others might just have some useful insight.
The age of the specialist (people who learn more and more about less and less, until they know absolutely everything about nothing) has proved to have some fundamental flaws. Three cheers for the generalists!
[1] http://www.scribd.com/doc/33655771/Economics-is-Hard
[2] For more on the weird and wonderful versions of Barbie that have graced the shelves over the years see http://shine.yahoo.com/channel/parenting/11-bad-barbie-ideas-1312923/?pg=8
[3] Atherya does have the sense to point out “Taken literally, I am almost certainly wrong.”
[4] John Ioannidis (2005) Why Most Published Research Findings Are False. PLoS Med 2(8): e124. doi:10.1371/journal.pmed.0020124
[5] Taleb (2007) The pseudo science hurting markets, Financial Times, 23 October 2007
I'm BAAAACK
Can't believe that it has been over two years since my last blog post! The good news (if you like my writing) is that I'll be posting the odd comment on this blog once again. As ever, I'll only post when I have something to say. The first entry will follow shortly.
Just for the record, I now work as memeber of the asset allocation team at GMO. However, the views expressed on this blog are entirely mine, and mine alone. They in no way reflect the views of GMO.
Just for the record, I now work as memeber of the asset allocation team at GMO. However, the views expressed on this blog are entirely mine, and mine alone. They in no way reflect the views of GMO.
Saturday, 12 January 2008
An update
Many thanks for visiting my blog. I have now returned to work full time with Soc Gen. I'm afraid I will not have the time to update this blog. My thoughts on behavioural finance and value investing will now be published by Soc Gen. In fact they have already published three - on simplicity, underperformance and action bias. If you are an institutional investor who would like access to my work please let me know.
Wishing you all the very best
James
Wishing you all the very best
James
Tuesday, 13 November 2007
The little book that makes you rich: what is really going on?
In my last post I questioned the long term relevance of some of the fundamental factors outlined in the Navellier's Little Book that makes you rich. However, I also noted that all eight of these factors only add up to a 30% weight in his final analysis. The other 70% is given to what Navellier calls his quantitative stock grade. He describes this as a measure of buying pressure amongst institutional investors.
However, he also tells us exactly what this measure actually is. On page 88 he says "In basic form, we divide a stock's alpha (the return independent of the stock market that typically comes from buying pressure) by its standard deviation. We measure this over a 52-week period".
The 'alpha' Navellier calculates comes from a simple CAPM model. However, as we show in Chapter 35 of Behavioural Investing, the simple CAPM model is deeply flawed. It just doesn't work. In fact in general there seems to be a negative relationship between beta and return, rather than a positive one.
Of course, Fama and French suggested a revised multifactor model of asset pricing - based on size, and price to book as well as the normal market factor. To help reduce the pricing errors in this model a momentum term was introduced by Cahart. A very recent modification has been proposed by Hirshleifer
. This adds a new factor to the equation of repurchases minus issuers (labeled UMO). This again reduces the significance of the alphas calculated from the FF4 model. In general the alphas become statistically insignificant under this five factor model.
So effectively, Navellier is running a semi reduced form model, not specifying which factors matter, but rather taking the alpha as a catch all term (which could be broken down into more understandable elements - such as size, value, issuance and momentum). However, Navellier demonstrates that these alphas have persistence. He estimates them over the past 52 weeks, and then uses them going forward.
To my mind this is consistent with recent work on style momentum. Chen and De Bondt have shown that style categories have a degree of persistence. They show that if you buy styles that have done well in the last year (in terms of size, price to book and dividend yield) they continue to do well over the next 12 months (but not beyond). The return achieved from a long short position based around this style momentum is around 7% p.a. using a 12 month holding period, and style past returns calculated over 12 months. In long only space a style momentum strategy generates a return of around 17% p.a. over the period 63-97. So Navellier's idea of alpha persistence certainly gets some support from this viewpoint.
Some final thoughts
I found much to agree with in Navellier's Little Book such as the over-reliance on stories, and the meeting with company management being a waste of time. His reliance on numbers based analysis echoes very much my own views on evidence based investing. However, ultimately I found the book couldn't stick with its own discipline. For instance, Navellier can't help but eulogize over the wonderful outlook for stocks that deliver out future. Despite his pronouncements that his eight factors and his quantitative grading system are really all you need to invest, he spends a considerable amount of time telling you to read the newspapers, whilst simultaneously ignoring the noise. Such overtly contradictory advice can do little but confuse the reader.
Personally I am not convinced that Navellier puts together a coherent defense of growth investing. But then again that won't surprise those of you who know me!
However, he also tells us exactly what this measure actually is. On page 88 he says "In basic form, we divide a stock's alpha (the return independent of the stock market that typically comes from buying pressure) by its standard deviation. We measure this over a 52-week period".
The 'alpha' Navellier calculates comes from a simple CAPM model. However, as we show in Chapter 35 of Behavioural Investing, the simple CAPM model is deeply flawed. It just doesn't work. In fact in general there seems to be a negative relationship between beta and return, rather than a positive one.
Of course, Fama and French suggested a revised multifactor model of asset pricing - based on size, and price to book as well as the normal market factor. To help reduce the pricing errors in this model a momentum term was introduced by Cahart. A very recent modification has been proposed by Hirshleifer
. This adds a new factor to the equation of repurchases minus issuers (labeled UMO). This again reduces the significance of the alphas calculated from the FF4 model. In general the alphas become statistically insignificant under this five factor model.
So effectively, Navellier is running a semi reduced form model, not specifying which factors matter, but rather taking the alpha as a catch all term (which could be broken down into more understandable elements - such as size, value, issuance and momentum). However, Navellier demonstrates that these alphas have persistence. He estimates them over the past 52 weeks, and then uses them going forward.
To my mind this is consistent with recent work on style momentum. Chen and De Bondt have shown that style categories have a degree of persistence. They show that if you buy styles that have done well in the last year (in terms of size, price to book and dividend yield) they continue to do well over the next 12 months (but not beyond). The return achieved from a long short position based around this style momentum is around 7% p.a. using a 12 month holding period, and style past returns calculated over 12 months. In long only space a style momentum strategy generates a return of around 17% p.a. over the period 63-97. So Navellier's idea of alpha persistence certainly gets some support from this viewpoint.
Some final thoughts
I found much to agree with in Navellier's Little Book such as the over-reliance on stories, and the meeting with company management being a waste of time. His reliance on numbers based analysis echoes very much my own views on evidence based investing. However, ultimately I found the book couldn't stick with its own discipline. For instance, Navellier can't help but eulogize over the wonderful outlook for stocks that deliver out future. Despite his pronouncements that his eight factors and his quantitative grading system are really all you need to invest, he spends a considerable amount of time telling you to read the newspapers, whilst simultaneously ignoring the noise. Such overtly contradictory advice can do little but confuse the reader.
Personally I am not convinced that Navellier puts together a coherent defense of growth investing. But then again that won't surprise those of you who know me!
Wednesday, 7 November 2007
The little book that makes you rich: A critical analysis of the fundamental factors
Firstly apologies for the recent lack of posts, I've been enjoying a sojourn visiting some of my family in New Zealand. I've recently been reading Louis Navellier's 'Little Book that makes you rich' subtitled "a proven market beating formula for growth investing".
I'm generally skeptical of the benefits of growth investing. All too often growth investing simply seems to be a cover for buying the latest fad or fashion in the investing world. So when I saw a book purporting to offer a numbers based approach (what I have called evidence based investing) to growth investing I was intrigued.
Navellier starts out by listing out his eight criteria for fundamental investing.
1. Earnings revisions
2. Earnings surprise
3. Sales growth
4. Operating margin growth
5. Cash flow to MV
6. Earnings growth
7. Earnings momentum
8. ROE
When I looked at this list I was somewhat surprised. Many of these factors struck me as odd. For instance, I have never come across a single paper claiming that sales growth had any kind of positive relationship with returns, nor ROE. Others such as earnings revisions and surprises were less shocking.
I decided to run a quick check on each of these factors, using a variety of sources ( I will run a full set of tests once I'm back at work, but for now I'll rely on others results). For each factor I tried to find the study with the longest history. The table below presents the results showing how much each factor added to a long only portfolio vs the market.
1. Earnings revisions 4.8% p.a
2. Earnings surprises 2.7% p.a.
3. Sales growth -13% p.a
4. Operating margin growth N/A
5. Cash flow to MV 4% p.a.
6. Earnings growth -2% p.a.
7. Earnings momentum 0% p.a.
8. ROE 0.8% p.a
In fairness to Navellier, he does note that the importance of each of these factors waxes and wanes over time. However, with a number of his factors appearing to add no value over a consistent time horizon, one must wonder what these fundamental variables bring to the party?
Interestingly, one of the best fundamental factors turns out to be a value factor! Although Navellier dresses up his use of cash flow as a growth variable, nothing can alter the fact that it is really a value variable. This is consistent with work that I have done which showed that value strategies did well within a growth universe (see Chapter 31 of Behavioural Investing).
It is also noteworthy that despite spending around two thirds of the little book of these variables, they only get a 30% weight in the final system....so what is the this little book really doing? I'll examine this in my next post.
I'm generally skeptical of the benefits of growth investing. All too often growth investing simply seems to be a cover for buying the latest fad or fashion in the investing world. So when I saw a book purporting to offer a numbers based approach (what I have called evidence based investing) to growth investing I was intrigued.
Navellier starts out by listing out his eight criteria for fundamental investing.
1. Earnings revisions
2. Earnings surprise
3. Sales growth
4. Operating margin growth
5. Cash flow to MV
6. Earnings growth
7. Earnings momentum
8. ROE
When I looked at this list I was somewhat surprised. Many of these factors struck me as odd. For instance, I have never come across a single paper claiming that sales growth had any kind of positive relationship with returns, nor ROE. Others such as earnings revisions and surprises were less shocking.
I decided to run a quick check on each of these factors, using a variety of sources ( I will run a full set of tests once I'm back at work, but for now I'll rely on others results). For each factor I tried to find the study with the longest history. The table below presents the results showing how much each factor added to a long only portfolio vs the market.
1. Earnings revisions 4.8% p.a
2. Earnings surprises 2.7% p.a.
3. Sales growth -13% p.a
4. Operating margin growth N/A
5. Cash flow to MV 4% p.a.
6. Earnings growth -2% p.a.
7. Earnings momentum 0% p.a.
8. ROE 0.8% p.a
In fairness to Navellier, he does note that the importance of each of these factors waxes and wanes over time. However, with a number of his factors appearing to add no value over a consistent time horizon, one must wonder what these fundamental variables bring to the party?
Interestingly, one of the best fundamental factors turns out to be a value factor! Although Navellier dresses up his use of cash flow as a growth variable, nothing can alter the fact that it is really a value variable. This is consistent with work that I have done which showed that value strategies did well within a growth universe (see Chapter 31 of Behavioural Investing).
It is also noteworthy that despite spending around two thirds of the little book of these variables, they only get a 30% weight in the final system....so what is the this little book really doing? I'll examine this in my next post.
Subscribe to:
Posts (Atom)
