'Analysts are rubbish'

[randomness]

From http://ftalphaville.ft.com/blog/2008/10/22/17316/montier-analysts-are-rubbish/:

"Seemingly everyone, on both sides of the Atlantic, is now taking about recession. Even [Bank of England governor] Mervyn King.

"So why, asks SocGen’s James Montier in his latest issue of Mind Matters, is the investment research industry still predicting earnings growth of between 12 and 15 per cent?

"He’s got a chart to illustrate that analysts are exceptionally good at one thing and one thing alone - telling you what has just happened."

Memo to analysts: management haven’t got a clue. They don’t know any better than the rest of us. One of the great delusions of our industry is that we all expect corporate managers to know more than we do. Whilst their knowledge may be deeper than ours, it certainly doesn’t translate into a better ability to forecast the future…

It is this optimism that is being drip-fed to analysts, who, seemingly incapable of any form of independent thought, are happy to regurgitate management’s mindless drivel to investors. This raises yet more questions as to why it is that we all employ so many analysts, if their major function is just to be a quasi-IR department for corporates.

Comments

[meepodeekin.livejournal.com]

That chart is hi-f*cking-larious. So true, too.

Of course, if analysts told us what had just happened with more info and insight than anyone else, they'd still be doing a useful job. Just not the job they think they're doing.

[holmes-iv.livejournal.com]

I concur 100%, on both points. But especially the first one. :-D

[drwex]

That's awesome.

[cerebralpaladin.livejournal.com]

One of the things I find interesting about that chart is that while there is clearly a good correlation between what happened ~1 year previously and what the analysts predict, it's not perfect. It makes me wonder if the differences represent useful information, or whether the differences are noise. If they do represent something, you might be able to get a more accurate prediction by essentially subtracting out the year ago portion and then applying some magnifier than by going directly off the analysts' predictions. Of course, it's fairly likely that the variations are basically just noise, so...

The other interesting thing is that the predictions lag by that much. You would think that with quarterly filings and so forth, they would mostly lag by a quarter. But they don't-- it's much closer to a year, more than a year in some cases.

Thanks for sharing.

[kessel.livejournal.com]

Holy crap, that chart just blew my mind.