Analysing the Analysts

Tweet

The idea for this blog came about on Friday 1st September after a recent new IPO, Ramsdens Holdings RFX, came out with a very positive Trading Update and said “we expect our interim and full year profit before tax to be significantly ahead of market expectations.”

RFX operates Pawn Broker shops but also has a Retail Jewellery bit, Gold Dealing and also they do Foreign Exchange which sort of differentiates them from other Pawn Broker Stocks like H&T HAT.

Anyway, as mostly is the case, the Trading Update came out at 7am but by around 9.30am I was reading on Tweets that a couple of Analysts/Brokers had already upgraded their Earnings Per Share (EPS) Forecasts to something like 16p in 2 years time. To be honest, the actual EPS Numbers are not the relevant point here - what struck me was just how fast this was to produce new Forecasts.

If you think about it, I was reading the Tweet at about 9.30am but by then the helpful people who were whizzing round this Information had already picked it up themselves from some source or other and of course this means that the ‘window’ within which the Analyst had done the Upgrade could easily have been well within 1.5 hours - which doesn’t strike me as very long at all.

This got me thinking about what an Analyst perhaps ought to do when they Upgrade a Forecast and also what impact such a swift reaction has upon the way in which they produce the Upgrade and the likely implications for us as Investors. Just thinking about this sensibly, I reckon an Individual Analyst should perhaps go through the following steps in an ideal world:

• Read the RNS Update several times to really understand the key facts within it and to get an idea of the rough impact on Forecast Numbers.
• Draw up a list of the Key Points from the RNS with a view to determining an Earnings affecting impact of each one.
• Carefully go through the List and make sensible judgements as to how the Revenues, Costs, Profits, Earnings Per Share etc. will be impacted (bear in mind that some aspects might have a negative impact on the Results).
• Call a meeting with other members of the Analyst Team (some of whom may have knowledge of the Company but most will probably have Companies of their own that they look after), where the RNS, its Key Points and their Impact on Earnings etc. are reviewed and tweaked if necessary.
• Contact the CEO and/or CFO of the Company and run the revised Numbers past him or her - I am not totally sure if this is allowed but it is clear that Companies often give Analysts and House Brokers in particular access to ‘Guidance’. This step could take place before or after the Team Meeting although it is appropriate that the Management Team has sight of these Upgraded (or Downgraded) Numbers before they are released to the Market. Any revisions as a result should be applied to the numbers.
• Obtain final approval from the Senior Management of the Brokerage prior to releasing the revised Numbers to the Market.

Anyway, I have never worked in a Brokerage or anything so this is pure guesswork but it strikes me these are the kinds of steps which ought to be undertaken if a realistic and agreed Revision to the Numbers is going to be produced.

So what on earth happened on the Morning of Friday 1st September when they managed to put out Revised Numbers for the Next 2 Years within maybe 1.5 hours? My hunch/understanding is that Analysts/Brokers have a hidden code where the use of the words “significantly ahead” means something like a 20% uplift or whatever and this is what the Analysts does - i.e. he or she just shoves 20% onto the existing EPS Numbers and that’s the job done. This is about as accurate and valid as tossing a coin I would suggest. In the case of RFX on Friday morning, I suspect the Analyst wanted to get down the PUB (which seems fair enough really !!).

In a similar vein, if the words “materially ahead” are used, this means a different percentage etc. (maybe 10%) and “broadly in line” means something else, which might even be a bit negative - it is all pretty hopeless if you ask me.

Bias in the Numbers
Bearing in mind what I have said above about how Analysts are probably interpreting these Trading Updates, my thoughts then rolled to what this would mean in terms of the Forecast Numbers - and the overwhelming conclusion was that 2 self-reinforcing Psychological predispositions would most likely come into play.

Firstly would be a natural Conservative Bias which is pretty normal in ’Spreadsheet Jockeys’ and is especially rife among the Accounting Profession - not that all Analysts come from such a background but there is no doubt they will have studied Accounting to some extent within their Education and a Conservative tendency is continually ingrained during such teachings. Secondly, ‘Anchoring Bias’ will have a profound impact where the prior Numbers before the Upgrade (or Downgrade) will weigh heavily upon the mind of the Analyst and mean that large movements away from that Anchor Position will be extremely unlikely. 

Both these effects will mean that the Analyst (in the case of an Upgrade) will tend to underestimate the Numbers that will be achieved in reality and if they do run their Revised Numbers past the Management of the Company, then they will also naturally err on the side of Conservatism because they know exactly how they can be punished if they set an Expectation which is too high and then they miss it. In extremis it will cost them their job (and it often does).

This effect will also come into play when the Management of the Company are deciding upon the wording for the Trading Update - it is far safer for them to use the word that suggests an Upgrade of 10% than to use a Word that is code for perhaps 15% or 20% and then to miss it. So there is a bias towards the low end all the way through the process.

This is also enhanced by simple ‘Rounding’ of numbers. Say for instance that Management thinks they will beat Expectations by 13%; due to simple Rounding and a slant to conservatism and some ‘Margin for Safety’, they will probably flag that they expect to beat by 10% - so again they are under forecasting.

What does this mean for our Stocks?

Uptrends and Upgrades
It is well know that Stocks which have a Share Price with strong Upwards Momentum and a nice Defined Uptrend tend to keep rising for perhaps much longer than people expect and that what appear to be very high valuations (a Forward P/E Ratio above 20 or so perhaps) often don’t matter so much because the Actual Numbers achieved come in far higher than expected which means that the Actual P/E Ratio is far less.

The way that Forecasts are Upgraded and the Bias I have talked about in the Section just above probably go some way to explain how not just the Share Price and the Business can have a building Momentum within them but also that the Forecast Upgrades can have a Momentum of their own and one Upgrade often leads to further Upgrades and so on.

In other words, the Conservatism of the Analyst Forecasts has a direct impact on the ability of the Share Price to keep rising when a Stock/Company is doing well. It is anecdotal, but I am certain I see this kind of Share Price and Upgrades behaviour all the time.

Downtrends and Downgrades
On the flipside, when a Stock has trouble and puts out a surprise Profit Warning, we often get a scenario where more Profit Warnings follow and a Share Price that gets into a Downtrend Channel just keeps falling until finally a Nadir is reached and at some point a Positive Trading Update comes out and the Shares can start to truly bottom out and begin to move up again. Often this happens not long after a New CEO and/or Chairman has been appointed and a ‘Kitchen Sink’ Trading Update has come out which has finally got every bit of Bad News out and in the open - easy for a new CEO to do but very difficult for a CEO who is tainted by the failed decisions of the past.

Think about this along similar lines to what I have discussed above about how the Analysts come up with their Revised Numbers - but now it is a process in reverse. In this case, the chances are that due to Conservatism, Anchoring and now a new Psychological Bias towards ‘Wishful Thinking’ (and I can imagine this occurs whilst the CEO has his/her fingers crossed behind their back), any new Forecasts that are produced will tend to be too light and not fully represent the true impact of how bad Trading actually is.

This means that when the next Trading Update comes out, yet again the Market is disappointed because the Actual Numbers will be worse than expected and another Downgrade gets done which again will feed the Share Price Momentum downwards. This process will continue until the Management is changed and the Numbers are properly Kitchen Sinked and we can start on the Upgrade Cycle (including of the Share Price, the Newsflow and Broker Upgrades) when the biases that apply to Uptrends will come into play.

Anyway, I hope that is an unusual take on something we experience every day in the Markets and will help you think differently about what is really going on with a Stock and where abouts the Share Price really is in the never-ending Cycle of Uptrend, Sideways, Downtrend, Sideways, Uptrend, Sideways etc. etc.

Organisational Forecasting Bias
I thought of this bit after I had done an Initial Draft of this Blog and thought it was worth shoving in because it is sort of related and is yet another example of how Human Psychology and in fact Organisational Behaviour (oh no, memories of University Courses - sorry @InvestingMartin !!) can lead to big Forecasting Errors.

A great example which has stayed with me for many many years is something that happened to ‘Big Blue’ IBM a long time back. I am sure most Readers know that IBM is a huge Computing Bureaucracy and it was this very structure which caused a massive Profit Warning that came as a total surprise to the Markets at the time - especially because IBM had been seen as a Company that could really do no wrong. I don’t know the exact date but at the time I think it was the biggest Corporate Loss in history.

The essence of the problems lay in the way the Company was structured and its culture which meant that giving someone above bad news was totally not acceptable. What happened was that down at the Ground Floor of the Business, People knew that Products weren’t selling well and Results would be very poor - so they communicated this up to their Bosses and this was deeply upsetting to them.

As a result, and because of this culture of not giving bad news, the Bosses then passed the Forecasts for Sales and Profits up to the Next Level of Management but they added in some ‘Judgement’ so that the Forecasts didn’t look quite as bad as what the Ground Floor Troops had actually flagged up. The Next Level of Bosses was not too pleased with this but of course they had to pass the Information up to their Bosses at the next level (you should be able to spot already that a Hierarchical and certainly not ‘Flat’ Structure was a major problem here with the Management very divorced from the true reality on the ground) and they did exactly the same thing so that they added in a bit of Judgement which just made the Forecasts even more padded and unrealistic.

This went on and on until once the Information reached Board Level of IBM, the Forecasts actually looked very good but with IBM’s history of achievement, it is probable that even they added some more on - so the whole situation became very surreal indeed.

Of course we all know what happened next and there was utter terror once the Actual Results were known - and Heads certainly rolled. Needless to say this kind of thing is a huge problem with vast FTSE100 Conglomerate type Businesses but it is also an issue for all Companies. Fortunately with smaller Companies there is probably a bit less of this as the Directors are a bit closer to the actual Coal Face.

Bonus Material
By a Simple Twist of Fate, this Week’s Investors Chronicle (1st September - - 7th September 2017 Edition, p.23) contains a bit by Chris Dillow entitled ‘Why momentum works’. Within his article he talks about some research by a Professor Titman (stop giggling at the back, this is not a name I made up myself) which blames it on investor’s overconfidence:

“This, he says, takes two forms. One is that (some) investors have more faith in their own beliefs than they do in others’. When good news about a stock arrives, overconfident investors who were sceptical about the firm’s merits sell as others buy. This causes the share to under react initially to that good news and hence to rise later. This is consistent with post-earnings announcement drift - the tendency for shares to rise in the days following good results in violation of the efficient market theory’s prediction that such news should be immediately embodied in prices.”

“The second aspect is that traders who come late to a stock exaggerate their own knowledge and so believe a stock is under priced even if it has risen a lot recently. This is likely to be the case if a share’s rise causes a company to attract the attention of investors who had previously ignored it.”

The above bits make some sense although I am not sure I agree entirely with these comments but I have thrown them in for Readers to think about. However, I do think it is interesting with regard to how Overconfidence can be an element with regards to Selling - I hadn’t thought of it this way before (I have always felt that Overconfidence was more associated with Buying) but perhaps there is some sense to it - particularly because it seems to be a major failing of most Investors that we sell good Stocks too early. There is a school of thought along the lines of “If in doubt, do nothing” and perhaps this kind of relaxed approach to letting Stocks run is the best way in truth.

Broker/Analyst Targets
While we are on the subject, I just wanted to throw a small bit in here about how I take pretty much no notice whatsoever of Broker Target Prices - I’m not remotely interested.

What I am after from Brokers is the Consensus Forecasts for EPS (Earnings Per Share) and anything else I can get like Revenue Forecasts, Dividends, Debt etc. In addition, if I do get to see a Broker Note they are often good for improving my understanding of the Business Model and you sometimes get insights into things like Operational Gearing and suchlike which can help. In other words the Numbers Forecasts and the Text are a million times more useful than their Targets, which in general I find to be far too high anyway (note, this is especially the case on the more Junky AIM Stocks where the Targets really are extremely dangerous to the unwary).

Conclusion
However we cut it, there are clearly very strong Momentum effects in Share Prices which we would be totally daft to ignore - both on the upside and on the downside. It is my experience that Good News tends to take a long time to slowly ‘bake’ into Share Prices and despite Shares falling fast and sharp on a Profit Warning etc., it is often the case that a Share in a Downwards Spiral will make new Lows after an initial bounce and buying Stocks which have put out Warnings can be fraught with trouble and perhaps there is no need for it anyway.

Due to these Momentum Effects and this likelihood for Analysts to under predict, selling Stocks in strong Uptrends that are continually putting out Good News is quite probably a mistake. It is very likely that they will beat expectations and what appears to be an expensive Stock, may actually be a lot cheaper that it seemed. As in my own mantra, “If you really must sell, just do a TopChop”.

There is also a clear implication that when looking to Buy a Stock we shouldn’t necessarily be put off a Stock on a High P/E Rating if it is in a Strong Uptrend and it is experiencing an Upgrade Cycle. I am not a user of Stockopedia but I understand they do put a lot of emphasis on Stocks which have been getting recent Broker Upgrades as having the potential to outperform.

Thinking about this Analyst Bias, I wonder if any Analysts themselves really consider Bias in their Forecasting? Maybe they did some sort of psychology stuff at University or whatever but it would not surprised me if this is a subject which is entirely ignored. Again, Human Nature comes into play here - “our Forecasts are highly accurate, why would we need to think about Bias?”

Cheers, WD.