We document widespread ex post changes to the historical contents of the I/B/E/S analyst stock recommendations database. Across a sequence of seven downloads of the entire I/B/E/S recommendations database, obtained between 2000 and 2007, we find that between 6,594 (1.6%) and 97,579 (21.7%) of matched observations are different from one download to the next. The changes, which include alterations of recommendation levels, additions and deletions of records, and removal of analyst names, are non-random in nature: They cluster by analyst reputation, brokerage firm size and status, and recommendation boldness. The changes have a large and significant impact on the classification of trading signals and back-tests of three stylized facts: The profitability of trading signals, the profitability of changes in consensus recommendations, and persistence in individual analyst stock-picking ability.
Monday, 9 November 2009
Airbrushing analyst history
Hat-tip to Aled for alerting me to this rather interesting paper. Here's the abstract:
Sunday, 8 November 2009
Detective dissonance
A funny thing happened last weekend linked to this question of cognitive dissonance that I'm interested in. We ended up watching one of those ITV adaptations of an Agatha Christie novel with David Suchet as Poirot. I say ended up watching, actually it's become a regular fixture (hey, we're new parents, our social life is not what it was). This particular adaptation was of One, Two Buckle My Shoe which, like quite a few Agatha Christie plots, involves some character not being who they appear to be. Basically, early on we are introduced to the character of Mabelle Sainsbury Seale. Later, one of the other characters assumes her identity in order to help carry out a murder.
The thing is that in the TV adaption it should have been obvious that this is what happened (I haven't read the book so I don't know if it is also the case in the original). You see the actress playing the 'real' Mabelle a couple of times. Then later you also see the fake Mabelle. The strange thing was that when I watched the programme at one level I clocked this. In a scene where Poirot (who never sees the real Mabelle alive) bumps into the fake Mabelle, I thought to myself 'that doesn't look like the same actress'. And in fact they were really quite different. However, because at that time the course of the plot wasn't clear, and because she kept being referred to by other characters as if she were the real Mabelle, I started to convince myself that actually maybe it had been the same actress.
This, I humbly submit, was an attempt to reduce the dissonance between two ideas - 1. "that is a different actress" and 2. "everyone is referring to her as if it is the same character". In attempting to do address this I (subconsciously) decided to downplay the evidence of my own eyes in favour of what others had said.
Subsequently it became clear that were indeed a real Mabelle and a fake one, and all along I had an uneasy feeling about the identity of the person that Poirot met. So the dissonance was never eliminated. But I did go a long way to ignoring what I had actually seen in favour of what others had said about the same event. Now it strikes me that if I can pull this kind of trick on my own mind in respect of the entirely trivial matter of watching a TV programme, then this may affect my decisions much more strongly in others areas (especially where there isn't any physical evidence). And actually when I think about it I can see attempts to reduce dissonance in lots of areas of my life.
Equally I think I see it a bit in the bit of the world I inhabit. There is IMO a tendency to take what a lot of organisations say at face value (particularly if you are paying them a large fee). If, for example, there is evidence to suggest that Organisation X's behaviour isn't quite in line with the way that they promote themselves to the market, I suspect the tendency on the part of those that employ that organisation will be to downplay the evidence. And indeed some of the conversations I have had with trustees justifying decisions made by their fund manager would seem to bear this out.
There is no answer to this. And, as Tavris and Aronson point out, actually the reduction of dissonance through self-justification (like many heuristics) enables us to make decisions quickly and stick with them. This is not unimportant. But equally we need to be constantly aware that attempts to reduce dissonance may result in the suppression of useful information that might be used to solve difficult problems, be they a fictional murder, or ineffective organisational relationships.
The thing is that in the TV adaption it should have been obvious that this is what happened (I haven't read the book so I don't know if it is also the case in the original). You see the actress playing the 'real' Mabelle a couple of times. Then later you also see the fake Mabelle. The strange thing was that when I watched the programme at one level I clocked this. In a scene where Poirot (who never sees the real Mabelle alive) bumps into the fake Mabelle, I thought to myself 'that doesn't look like the same actress'. And in fact they were really quite different. However, because at that time the course of the plot wasn't clear, and because she kept being referred to by other characters as if she were the real Mabelle, I started to convince myself that actually maybe it had been the same actress.
This, I humbly submit, was an attempt to reduce the dissonance between two ideas - 1. "that is a different actress" and 2. "everyone is referring to her as if it is the same character". In attempting to do address this I (subconsciously) decided to downplay the evidence of my own eyes in favour of what others had said.
Subsequently it became clear that were indeed a real Mabelle and a fake one, and all along I had an uneasy feeling about the identity of the person that Poirot met. So the dissonance was never eliminated. But I did go a long way to ignoring what I had actually seen in favour of what others had said about the same event. Now it strikes me that if I can pull this kind of trick on my own mind in respect of the entirely trivial matter of watching a TV programme, then this may affect my decisions much more strongly in others areas (especially where there isn't any physical evidence). And actually when I think about it I can see attempts to reduce dissonance in lots of areas of my life.
Equally I think I see it a bit in the bit of the world I inhabit. There is IMO a tendency to take what a lot of organisations say at face value (particularly if you are paying them a large fee). If, for example, there is evidence to suggest that Organisation X's behaviour isn't quite in line with the way that they promote themselves to the market, I suspect the tendency on the part of those that employ that organisation will be to downplay the evidence. And indeed some of the conversations I have had with trustees justifying decisions made by their fund manager would seem to bear this out.
There is no answer to this. And, as Tavris and Aronson point out, actually the reduction of dissonance through self-justification (like many heuristics) enables us to make decisions quickly and stick with them. This is not unimportant. But equally we need to be constantly aware that attempts to reduce dissonance may result in the suppression of useful information that might be used to solve difficult problems, be they a fictional murder, or ineffective organisational relationships.
Saturday, 7 November 2009
Bela Lugosi's Red
Interesting factoid that I discovered via a documentary today - in addition to being a trade union activist, Bela Lugosi was actually involved with the 1919 communist revolution in Hungary, led by another Bela - Bela Kun. He was heavily involved in the Hungarian actors' union and it appears he was actually had a minor role in Kun's administration. This was a problem for him later on when the FBI kept tabs on him, something that appears to have damaged his career.
Thursday, 5 November 2009
Data disappears...
Most people have read about the unfortunate disbanding of the well-regarded responsible investment team at Insight Investment. Having had a look at the Insight website today it appears that the ‘Responsible Investment’ section, which had quite a lot of useful documents in it, has been removed or made inaccessible. This also means that you can’t access their voting and engagement records, which were presented in one of the best systems I have come across in my voting geekery.
So let me mount my hobby-horse another time and take it for a quick ride. Doesn’t this just demonstrate the problem of a voluntary system of disclosure, where the data is only provided at an asset manager’s discretion? Insight started disclosing its voting back in 2004, and it was a big asset manager, so there was a lot of data there. Now it is gone.
Because there is no central repository for voting data, where we could access voting decisions from previous seasons regardless of whether the manager had been taken over, unless someone at decides to reverse this decision I presume Insight's voting record is gone for good. I think it's unlikely that SWIP (who took the equities bit of Insight) will suddenly decide to make it all public again, for the simple reason that they don't disclose any data themselves.
In contrast, if we had a mandatory system, where managers had to make an annual report on how al their votes have been cast, and file it with an organisation like the FRC, this problem would not emerge.
So let me mount my hobby-horse another time and take it for a quick ride. Doesn’t this just demonstrate the problem of a voluntary system of disclosure, where the data is only provided at an asset manager’s discretion? Insight started disclosing its voting back in 2004, and it was a big asset manager, so there was a lot of data there. Now it is gone.
Because there is no central repository for voting data, where we could access voting decisions from previous seasons regardless of whether the manager had been taken over, unless someone at decides to reverse this decision I presume Insight's voting record is gone for good. I think it's unlikely that SWIP (who took the equities bit of Insight) will suddenly decide to make it all public again, for the simple reason that they don't disclose any data themselves.
In contrast, if we had a mandatory system, where managers had to make an annual report on how al their votes have been cast, and file it with an organisation like the FRC, this problem would not emerge.
Wednesday, 4 November 2009
Select committee shuffle
I've spent a few hours at the Treasury committee hearings this week - yesterday was Sir David Walker, today UKFI followed by Paul Myners. I wasn't able to attend the 'show trials' in February when the likes of Fred Goodwin and Andy Hornby gave evidence, and these sessions were much more sparsely attended.
Nonetheless there was some interesting stuff in there. I thought some of what Walker said was quite encouraging. In particular he moved on to talking about shareholder engagement without prompting from the committee. And judging from some of his comments (ie scepticism about engagement in the recent past) he appears to have some idea of the very real challenges in getting investors to take these things seriously. There was no-one there except us from the investor world, which I was surprised about.
Today I was really impressed by John Kingman. His answers were (mostly) very fluent and he came across as very much on top of his area. Which is rather important since he has been in charge of our investments in the banks. He definitely seemed to have a clear idea of who had responsibility for what in the HMT-UKFI-banks interface. UKFI did take a bit of flak for not revealing the bonus pot for 2008 for RBS. And Michael Fallon (rightly) pulled Kingman up at one point for "mandarin-speak" but all in all it was a polished performance. There was some interesting discussion too about the relative weight UKFI puts on representing the taxpayer's interest as shareholder and as bank customer.
Next up came Paul Myners, though we ducked out a bit early when they took a 10 minute break in the meeting as this session it wasn't on a topic of direct relevance to us. Given their recent encounters you might have expected this to be quite a fractious session, but actually what we saw of it was quite good-natured. That said there did seem to be a bit of a crackle in the air between Myners and Fallon.
Finally, the committee itself. I'm quite impressed by John McFall and he was quite punchy today in particular. And generally I had less of a sense of them missing the right questions to ask, whilst asking the 'populist' ones, though obviously I would have asked some different ones myself. It's democracy - a bit messy, doesn't work brilliantly sometimes, but it seems to do the job we want/need doing.
Nonetheless there was some interesting stuff in there. I thought some of what Walker said was quite encouraging. In particular he moved on to talking about shareholder engagement without prompting from the committee. And judging from some of his comments (ie scepticism about engagement in the recent past) he appears to have some idea of the very real challenges in getting investors to take these things seriously. There was no-one there except us from the investor world, which I was surprised about.
Today I was really impressed by John Kingman. His answers were (mostly) very fluent and he came across as very much on top of his area. Which is rather important since he has been in charge of our investments in the banks. He definitely seemed to have a clear idea of who had responsibility for what in the HMT-UKFI-banks interface. UKFI did take a bit of flak for not revealing the bonus pot for 2008 for RBS. And Michael Fallon (rightly) pulled Kingman up at one point for "mandarin-speak" but all in all it was a polished performance. There was some interesting discussion too about the relative weight UKFI puts on representing the taxpayer's interest as shareholder and as bank customer.
Next up came Paul Myners, though we ducked out a bit early when they took a 10 minute break in the meeting as this session it wasn't on a topic of direct relevance to us. Given their recent encounters you might have expected this to be quite a fractious session, but actually what we saw of it was quite good-natured. That said there did seem to be a bit of a crackle in the air between Myners and Fallon.
Finally, the committee itself. I'm quite impressed by John McFall and he was quite punchy today in particular. And generally I had less of a sense of them missing the right questions to ask, whilst asking the 'populist' ones, though obviously I would have asked some different ones myself. It's democracy - a bit messy, doesn't work brilliantly sometimes, but it seems to do the job we want/need doing.
Tuesday, 3 November 2009
Remuneration and electric shocks
I've just started reading Mistakes were made (but not by me) which is all about self-justification, typically arising from cognitive dissonance, and the way it distorts decision-making. Obviously it immediately made me think of the way the former heads of RBS and HBOS defended themselves in front of the Treasury select committee. But it also made me think about shareholders' justification for their recent activity.
Early on the book talks about the famous Milgram experiment (the one where subjects thought they were giving shocks to a person in another room when they got questions wrong). The experiment is principally known because it appeared to show how people defer to authority, but the book makes the point that it also tells us something about self-justification:
I think at each stage we had a similar opportunity to challenge. But then the company putting forward a given policy - the authority figure (and unfortunately some corp gov people swoon a bit when faced with a real-life executive) - provided a justification. Our industry is different - we need to pay top dollar to recruit and keep the talent. So the investors let it slide. Next time the company comes back and says we need to pay more to stop the talent moving into private equity. Ummm.... ok then. And so on.
At each point most shareholders deferred, and in doing so will have justified their decision to themselves, and become more wedded to it. Therefore we reached the stage prior to the crisis where very few companies were facing serious shareholder pressure (no remuneration reports were defeated in 2008 remember). By that point, having let so much go past them, it must have been difficult to justify actually challenging a company that was only a bit worse than the others.
Early on the book talks about the famous Milgram experiment (the one where subjects thought they were giving shocks to a person in another room when they got questions wrong). The experiment is principally known because it appeared to show how people defer to authority, but the book makes the point that it also tells us something about self-justification:
[T]wo thirds of [the subjects] go all the way to the maximum level they believe is dangerous. They do this by justifying each step as they went along. This small shock doesn't hurt; 20 isn't much worse than 10; I've given 20, why not 30? As they justified each step they committed themselves further. By the time people were administering what they believed were strong shocks, most found it difficult to justify a sudden decision to quit.This immediately made me think about the way that shareholders have effectively legitimised high levels of executive remuneration. I have no doubt that many (most?) people working in corporate governance at asset managers believe, looking back, that things have got out of control, but how did we get here?
I think at each stage we had a similar opportunity to challenge. But then the company putting forward a given policy - the authority figure (and unfortunately some corp gov people swoon a bit when faced with a real-life executive) - provided a justification. Our industry is different - we need to pay top dollar to recruit and keep the talent. So the investors let it slide. Next time the company comes back and says we need to pay more to stop the talent moving into private equity. Ummm.... ok then. And so on.
At each point most shareholders deferred, and in doing so will have justified their decision to themselves, and become more wedded to it. Therefore we reached the stage prior to the crisis where very few companies were facing serious shareholder pressure (no remuneration reports were defeated in 2008 remember). By that point, having let so much go past them, it must have been difficult to justify actually challenging a company that was only a bit worse than the others.
Subscribe to:
Posts (Atom)