How is Econometrics Applied in ER?
Hi,
I am very curious as to what linear regression techniques are used by equity research analysts and for what purpose. Is there a resource devoted to this topic?
In my experience of working on four separate teams on the sellside (and a stint at a LO AM), I have never seen anybody use linear regression in their analysis. Speaking more broadly on econometrics (and finance academia in general), the vast majority of it isn't used in practice.
The most useful things I learned from school that I was actually able to apply on the job were mostly topics covered in accounting. The only things I've ever used from the finance classes I took were literally simple DCFs where our cost of equity was more of a plug than something we rigorously calculated.
Thank you Braininajar for your very helpful insight.
If ER analysts don't use linear regression, how do they forecast sales units in DCF?
Thanks again.
Regression isn't used for that. The goal of linear regression is to draw a trendline over existing data. It doesn't actually predict new values. Standard econometrics doesn't deal with predicting values in general. You could make predictions (forecasting) based on financial econometrics with models like ARMA & ARIMA or with models in the comp sci department using different ML algos e.g. RNN, LSTM, GRU etc.
dickthesellsider got it basically completely right.
The only bit i'd add is that for a lot of large cap / mature companies, they issue pretty regular guidance and sometimes even give quarterly guidance. 95% of sellside analysts will just model the short / medium term to company guidance +/- a few percentage points to satisfy whatever narrative they're pushing (i.e. buy or sell rating). And for the out years, they'll basically do what dickthesellsider said and just model some arbitrary steady-state growth. Realistically, nobody is going to look at your out year assumptions unless you're way off of consensus.
Also, just in case you think i'm criticizing analysts for following guidance, i'm not. These companies have armies of FP&A analysts who have way more granular information than any external observer who crunch the numbers to help come up with whatever high-level guidance there is (e.g. we expect revenue growth of 3-5% in xyz segment). Unless you have a real fundamental view for why you think the company is overestimating or underestimating their prospects, you should almost never deviate from near term guidance simply because you're much more likely to be wrong.
I rarely see fundamental forecasts done using regression, but regression is used often enough in charts where they correlate valuation multiple with some fundamental metrics and use a high R^2 to justify explanatory power of a spurious relationship.
You see it in SaaS research reports at a lot of shops where they relate say, NTM growth rate with EV/rev multiple of all the pub SaaS names and draw a line between all the data points and say "look, 70% R^2, so if something is growing at 30%, it should trade around this rev multiple".
Forecasting growth / margin is not supposed to be that scientific, it's more like "look, historically this industry can do 3% volume growth and can increase price 2%, so revenue can go up at most 6% each year and you just use that during your forecast period, and the best company in the industry can do 30% EBITDA margin, so if you are bullish, you have to believe your company can converge that margin gap in a reasonable number of years. If your name has 5% EBITDA margin right now, is it realistic / has any company in the space ever able to get to 30% EBITDA margin in 5 years from 5%? No? Then you can't count on 30% EBITDA margin as part of your thesis."
Exercitationem optio sapiente possimus quia libero omnis et. Inventore tempore maiores voluptas non enim cumque.
Nulla et distinctio eligendi veniam et. Sed mollitia consequuntur in aut asperiores aperiam. Dolorem sed et iusto quaerat doloremque. Quidem natus natus corporis minima cum.
Temporibus repudiandae natus veritatis perferendis recusandae. Recusandae quam sint dignissimos saepe ipsam officia voluptatum. Ipsum quae et veritatis culpa qui sint.
Adipisci sunt eligendi soluta nostrum. Hic adipisci quasi ex aut rerum ullam porro. Corporis sunt sit inventore mollitia modi. Earum eos modi repellat voluptates nobis alias modi. Sapiente enim doloribus expedita reiciendis quibusdam consectetur. Modi ratione ullam repellendus ullam.
See All Comments - 100% Free
WSO depends on everyone being able to pitch in when they know something. Unlock with your email and get bonus: 6 financial modeling lessons free ($199 value)
or Unlock with your social account...