Hi all, I’m trying to find background information on the earnings estimates we utilize in p123.
In particular, question like:
- who gets to make estimates? what qualifies someone (or some firm) to make an estimate?
- how easy would it be to manipulate estimates, and is the “system” designed to protect against that?
I ask because estimates are a useful predictor, and at the same time it seems given our reliance on estimates it seems like getting a “ringer” in the estimate pool to either raise or lower the consensus estimate would be an effective way a stock manipulator to create either positive or negative earnings surprises - especially in companies with low # of estimates/analysts. I guess I’m trying to gauge the honest accuracy, or degree of possible manipulability.
In the end, I’m wanting to be able to assess to what degree I should weight earnings estimates based on this - especially in small companies with very few analysts.
If anyone has thoughts or good info or has links to discussions on this I’d appreciate. I’ve had difficulty finding anything useful with web searches so far. Thanks in advance,
We get the estimate aggregates fully calculated from S&P, and have no real information on the underlying numbers. In general, though, the people who get to submit estimates are professional analysts. They are usually on the sell side – meaning that the reports are published for use by other people to make investing decisions – rather than on the buy side – meaning that they are portfolio managers who are making the decisions for their own portfolios. (It’s not in the buy side’s interests to share their analysis for the most part.)
How easy are they to manipulate? An analysis report is, essentially, an opinion piece. Good luck trying to spot the difference between a differing outlook and an attempt at manipulation, at least without a whole lot more information than you’re going to have about the companies and the analysts. If you’re really worried about it, we have the standard deviation of the estimates as a data point. Presumably a low standard deviation is a stronger consensus.
Oh, and it’s a problem that essentially goes away with enough liquidity. Nobody does pump and dump with Apple.
Thanks for the info Paul. I actually do find the earnings estimates historically useful even for small companies with very low # of estimates (even 1 analyst). I guess I can’t help but be a little bit paranoid about it unless the analysts are rewarded for accuracy. With so many investors relying on estimates, especially in a quant context for small caps, it just seems like 1 analyst moving #s around could have a big impact on rankings. I guess that type of activity should degrade earnings estimate as a useful factor, so the fact that they do seem a useful factor even in the very small cap space is a good sign for reliability. thanks again 