Foreword
Opinions are like rear ends — everyone has one, and most of them smell funny. Analyst reports are just formal versions of opinions. Draw your own conclusions.
I want to start by noting that I am not a financial professional by training — I am a software engineer by training, and by trade. The following is based on my personal understanding, via some formal classes, but mostly self-taught.
If you find anything that you feel is incorrect, please feel free to leave a comment, and discuss your thoughts.
What report now?
The definition of an “analyst report” is a little loose — people have been talking about stocks pretty much since people have been trading stocks. Anyone who claims to be able to predict the movements of stocks will always get an audience.
To keep the discussions sane, when I say “analyst report”, I mean “analyst reports, valuation models and other stuff of this nature”. Essentially, any document (or video!) that tries to decipher the threads of Fate and give you an insight on what a stock’s price would be in the future.
Types of report
Analyst reports come in 3 main flavors – sell side reports, buy side reports and independent reports.
Sell side reports are the reports that banks, brokers, dealers, etc. generate. These are entities that generally are not investing in the stocks, but provide a service to help someone else (their clients) buy or sell stocks.
Buy side reports are the reports that hedge fund managers, private equity managers, endowment fund managers, private investors, etc. generate. These are entities that are investing in the stocks themselves, or are managing money for others who are investing.
The main difference between brokers/dealers and “money managers” in this, is that “money managers” (buy side) have “skin in the game” — if their recommendation do well, they tend to profit, and if it does not, they may lose money. Brokers/dealers (sell side), on the other hand, are generally just interested in encouraging trading activity — they collect a fee based on each trade, and have no further “skin in the game”, regardless of how the stock performs.
Independent reports are generated purely for the sake of the report. For example, independent research companies which generate reports, and then try to market and sell the reports themselves.
In terms of quality, independent reports tend to be the least biased, followed by buy side, followed by sell side.
Note: Everytime the market is moving rapidly, either up or down, there will be a rush of people trying to portray themselves as “gurus” of the stock market. Some of these people are legitimate — proper research operations with a team of researchers. Others are more fly-by-night operations with a single (or maybe husband+wife/family) operator, yet others are just thinly veiled buy side operations that are just touting their own stocks. The first may or may not produce good recommendations, but the latter two almost never so.
Why write these reports?
As hinted above, sell side reports are generally created as a means to encourage clients to trade more. For example, most brokerage firms will produce reports that provide basic information about a company, and provide historical charts of how the company’s stock price and various other metrics have performed. Some buy side reports also include projections or even recommendations on what stocks to buy and when.
The goal, ultimately, is to provide as much information as needed for the client to decide that they know enough to pull the trigger — to execute a trade. Remember, sell side earns their money from collecting fees (or spreads) when a trade happens, or by collecting fees for handling your account. If you don’t trade, and you don’t put money/assets with them, they don’t get paid.
Buy side reports, on the other hand, are generally private. They are generated as proprietary work products of large financial entities, or even your average investor! Many retail trader have some form of research report that they produce while trying to decide how to manage their money. This can be as simple as “TSLA to the moon!” scribbled on a piece of toilet paper, or as detailed as a spreadsheet with line by line breakdown of a company’s quarterly reports.
The buy side reports that generally make it to the public, usually are published with a single focus — to convince the rest of the world that they are right, and that the rest of the world should follow them in that trade.
Independent reports, finally, are usually made for sale. They tend to be more neutral in tone, and often, the goal of the report is to sell the report itself. For example, Morningstar, Motley Fool, Benzinga, Seeking Alpha all provide independently sourced reports for sale. (1)
Reading an analyst report
When you read an analyst report, you should keep in mind the main objective of the research author, as well as their competencies. As this blog clearly shows, anyone with a keyboard can put together a post. Whether that post is worth reading, is an entirely different matter!
For the most part, analyst reports are fine — they may be wrong (or right!), but they are “fine”. Which is to say — analyst reports are not always right. If you read a report in full, including all the little size 1 font wordings and maybe press the author for proper disclaimers/assumptions, you’ll quickly realize one thing:
All reports have a list of assumptions/caveats, that taken in full, will read something along the lines of, “This report is correct, assuming it is correct. It may also be wrong. Don’t sue us.”
Analyst reports are not meant to be crystal balls — they are not meant to be predictive. For the most part, they are meant to be persuasive. i.e: Given a set of assumptions, then one possibility is that “this” will happen, and you should believe me, because <reason>.
In many cases, the assumptions are simply “assuming what we saw in the past N months repeat in the next N months”. Which is “fine” — it’s a reasonable prior given no additional information, but it is not “right”, nor is it “predictive”.
When you read an analyst report, don’t just skip to the last line that says “stock X is worth $Y”. Because that line is, literally, the most useless line in the whole report.
That line bakes in the biases, prejudices and, frankly in many cases, dumb-posterior assumptions made by the author, along with whatever number/fact fudging they care to put in. Instead, read through the assumptions, and see if they make any sense. You need some amount of critical thinking, some background on the macro and micro environments, and potentially some research of your own.
Once you’re done with the assumptions, look at the model the author is building. There are many valuation models, but all of them have pros and cons. More importantly, not all models apply to all companies. For example, P/FCF is a very useful model for REITs, because of their tax structure, but P/E is completely useless (because a large part of “E” is reduced by depreciation, which isn’t a real cost for most real estate properties) (2).
Once you’ve done the above, there are a few ways to react to an analyst report:
- Read more analyst reports.
- Adjust their numbers that were based on wonky assumptions and/or model.
- Assign a probability for each report to become true, based on what you understand about the macro/micro environment.
- Then take a probability weighted average of all the adjusted results and use that result.
For example, after reading 3 reports, and adjusting each for obvious errors, you get these predictions:
Report 1: Stock @ $100
Report 2: Stock @ $90
Report 3: Stock @ $50
You give these reports the following probabilities of becoming true:
Report 1: 50%
Report 2: 40%
Report 3: 10%
And so, the weighted average is (100*0.5) + (90*0.4) + (50*0.1) = $91
- Read more analyst reports.
- Filter out those that are just plain batpoop crazy.
- Of the rest, look at the inputs they use, and for each input, consider a reasonable conservative estimate across all reports (you can use the most conservative, or the 25%-tile or whatever, depending on how risk-averse you are).
- Then recompute based on these numbers.
For example, if you filter down to 3 reports that are reasonable, and all of these have stock price models based on some estimate of future sales and future production costs, then you can either take the median (or 25%-ile, or average, or whatever) estimate for each of future sales/costs.
Plug these blended estimates into the model, and arrive at your own estimate for the stock price.
- Read more analyst reports.
- Use the reports to get a feel of what people “on the street” are thinking, because while a single report is probably noise, a bunch of them together may show a useful trend.
- Build your own model.
- Read the report as a work of fiction, just like Harry Potter. If you enjoy it, great. If not, maybe try Judy Moody instead.
- Roll your eyes at yet-another-crazy-analyst-report, say something nice but vague so that whoever showed you the report, and is eagerly hopping up and down telling you about this “hot new opportunity” that is “sure to go to the moon”, will just leave you alone.
- Start a thread in an obscure forum in a private company/blog, trying to explain that analyst reports are not meant to be prophetic, nor are they the threads of the Fates. And pray that enough will understand enough that they stop throwing money at terrible ideas based on even more terrible ideas.
Footnotes
- This is not a recommendation nor endorsement for any of these services, or the quality of their reports. Also, note that some research branded as “independent” may have ulterior motives, such as illegal pump and dump schemes, trying to “talk the author’s book”, etc.
- See the “How to value a company” series of posts for more details on valuation models:
- How to value a company – income statement
- How to value a company – balance sheet
- How to value a company – cash flow statement [coming soon]
Janky Thoughts 