Calling Quarters (forecasting earnings)
Hello all,
For the HF analysts who are shorter term focused in nature - I wanted to know what steps do you guys all take to call quarters/predict earnings accurately. I am someone who has no real clue and hide behind the "longer term investment horizon". I assume Multi-Manager L/S equity pod analysts are particularly good at this. I would love steps/advice on methodology to be able to get better at predicting short term earnings.
Thank you all for your help in advance.
Also curious. Is it just a lot of deep diligence into the revenue / cost build?
adding to this question, I'd be also curious how deep models for each ticker analysts usually build out?
Really depends on the company, and the industry. I don't think anyone is *that* good at it, but sometimes a bit of work makes it clear that sell-side estimates are improbable. Here's my toolset:
Revenue
Gross Margin
SG&A
So that's a high-level view of how you'd model earnings. But I think earnings are often the wrong metric to focus on. In so many industries, the most important KPI (because it's most relevant for longer-term growth) is the most forward-looking indicator of revenue growth: bookings/billings for software companies, backlog for industrial companies, comps for retailers, etc. Almost no company does well when its forward-looking revenue indicator is weak, no matter how much it beats on earnings. And even earnings misses don't matter that much if there are signs that revenue is accelerating. The market knows it's not hard to restore margins when revenues are healthy, but it's very hard to sustain margins when revenues are weak.
good stuff, thanks a lot!
Great post, I want to emphasize the comment about how most models forecast SG&A as a % of sales. Certainly if this % of sales measure is fairly consistent over time for a growing company (asset light people-based businesses) there may be some efficacy in taking that approach, but for most companies modeling as a % of sales will cause one to miss the operating leverage a growing company benefits from over time, and it is surprisingly common. On the flip side of that, you see models missing the same decremental margins on the downside, especially in this environment, for the same reason.
Is this where incremental / decrement margin needs to be incorporated?
This is a great summary and aligns with how I've been taught / my team does it. Since everyone seems to be doing pretty much the same thing, where does your variant view come from?
Good question. In recent years, I don't think there has been that much edge in just doing shorter-term earnings analysis. I think there was more alpha available a few years ago, but capital has flooded into the multi-manager space and eliminated a lot of the easy (easier) money. The way I try to adapt my own process is to really dial up my focus on the question of, "is this earnings result, or other catalyst, of significance to the long-term trajectory of the business." I put less focus onto getting the earnings precisely right, and more on getting a better understanding of what management, and/or a smart bull see as the positive trajectory for the business over the next 3-5 years, vs. what management fears, or a smart bear expects, for a negative outcome. Beats that prove true long-term bulls/bears right get you paid far more than beats on short-term issues. I'm fortunate to be in a seat where I can approach things a bit differently.
Yea I would emphasize an increasing focus on KPI’s that are seen as “leading indicators”. At least in my industry, stocks around earnings tend to trade more around the reported leading KPIs than the actual financials. More often than not, modeling revenue/EPS from these KPIs is pretty straight forward so the greatest point of sensitivity in models tends to be these KPIs.
Yep. Paypal 1Q20 earnings is an example of operating KPIs being more important than revenue/earnings. At this point, current revenue/earnings aren't so influential on same-day post-ER moves as the updated management guidance/outlook.
Would also ask: is it likely AI/algo-trading might cripple part of the market strategy based on it detecting patterns in the trading data unobservable by humans (where such patterns are correlated with KPIs)?
And when do you put on the position once you determine that it is will be a beat / miss? I don't usually trades quarters, but I recently shorted a name in anticipation of a miss using CC data.
I put it on around 1-2 weeks prior to earnings but the stock went up during that time and then, on the day of, it missed and fell by less than it went up. So I took a L.
Do you guys try to wait until the last possible minute to put on the trade to avoid having interim vol erase any potential gains from the actual miss/beat?
You're asking the wrong question. How many hundreds of people similar to you have access to the same CC data? And if they're seeing the same data you are, do you really have an edge?
> Do you guys try to wait until the last possible minute to put on the trade to avoid having interim vol erase any potential gains from the actual miss/beat?
You can trade whichever timeframe you have an edge for. Like, if you have an edge to trade earnings 2 weeks before the announcement with CC data -- do it. But yes, the closer to the announcement, the lesser the macro impact on the outcome, but the more people contribute to efficiency.
How do you identify and get in touch with suppliers? Why are suppliers willing to talk to you?
As for identifying suppliers, Bloomberg has a decent database, built on public company disclosure (companies often disclose their largest suppliers and customers, which can be used to find some of the larger ones). Some companies list their major customers on their websites, which makes it easy. Sometimes GLG or other expert networks have done the hard work for you already, so you can just ask them.
Some suppliers are just public companies, so they'll chat because any public company will; most of your questions have to be about their business, but you can slip in a few about their customers. As long as the call is mostly about their own company, they won't mind (and the behavior of their customers is relevant for them, too!). With respect to private suppliers, many executives of private suppliers are on GLG and similar networks, so arranging a call is simple.
Depending on how good (and proactive) your firm's compliance operation is, they can assist you in figuring out which suppliers are legal to talk to, and what you can legally ask them/they can tell you. Then you can reach out to them on LinkedIn directly to ask them for a paid consultation. This route is relatively onerous though, and not easy to do. Some firms do have in-house sourcing operations, however, to get those contacts lined up in a systematic fashion. I wouldn't attempt doing this without compliance input, but I think some people do it regardless.
Truly great stuff, I'm a LT (5yr horizon) investor myself and have long been wondering how you sub-1yr horizon guys do your analysis - lays it our very cleanly
On a separate note, do you feel this ST horizon (3mo<) is highly susceptible to AI / ML / quant? I've been hearing that this type of investing will most likely die out over next 2 decades while the LO stuff (3+yr) will likely remain relevant for much longer (although wouldn't be surprised if in ~4 decades that was automated away as well)
LOL are you kidding or just plain delusional? Citadel MLP etc keep prospering while every single LO keeps lagging the market and losing AUM share to passive ETFs.
Seriously, the lack of self awareness has me amazed.
Thank you for admitting that “long term investment horizon” is a stone to hide behind and not an intellectual high ground
Hmm don't think it's either. I'm fortunate to be able to look at 5+yr time horizons and wouldn't trade it for the world. I do think it's the best way to make money (time / stress / capital accretion weighted).
That said, there's plenty of ways to make money incl. quarterly horizons. The returns posted by Jack Woodruff are incredible. Don't think LT investing is a stone to hide behind just as it's not fair to condemn ST investing. Different strokes
That’s fair
so I've worked on much shorter time horizons but do admit there are positives to being able to drill deep into names and realize potential over 3-5yr horizon, but firms I have spoken to that do this have not made a convincing argument. Generally managers at these places run 60/40 and though arguably some have an amazing short book the long side almost is never differentiated and is tightly concentrated in 10-15 names that hugs the benchmark. Most have sharpe that is just above 1 (for comparison the benchmark is over above 1 so no outperformance).
As for AI, I don't think so. Quants have a lot more to fear than short term fundamental analysts.
As someone who tried, 5+ year time horizons don't work in public markets and a small part of why I ended up exiting the HF space.
People obviously have a hard time predicting what happens and that becomes exponentially harder and harder the further you go out. Thousands of factors regarding what the market does, what the company does, what competitors do, what management does, etc. The big problem is the famous John Keyne quote, you'll get shit on for years before it works out.
Partner Fund and countless others couldn't make it work. Ignoring the above, the main issue is that you are a minority investor. Not even management can predict 5 years out and they have their hands on the wheel. Using the same analogy, that's like if you were in the back seat of a car and you're trying to predict where both the driver and the road ends up by the time the tank is empty. To to the above rude poster's point, it's why short-term multi-managers are the only ones doing well because their process does work as minority shareholders (citadel, millennium, point 72) because you have to jump on your only advantage which is immediate liquidity. Edit: Oh and alternative data obviously helps here, sure it's somewhat a weapons race but credit card geolocation blah blah can still help savvy and creative investors on short term catalysts & quarters.
If you're going to limit short-term trading and go for long-term horizons, you're handicapping your only strength. Might as well go full illiquid and own entire companies / run them yourself. As you probably already know, that's called private equity.
I'm surprised no one has yet raised the point that calling a "beat" right might still result in losing money (stock goes down) if it's a well-expected beat and all the HFs are already positioned that way.
Getting the earnings forecast right is only half the battle. Determining how the stock will react matters just as much and is arguably even harder.
Esp in the era where so many bots are involved in the market
I think it's a style thing. Running your own PA will probably give you a good idea what types of situations you excel in. Very few people can call the quarters consistently and the ones that I've seen doing it (or say they do) in my observation are focusing more on the market/trade structure rather than fundamentals of how the quarter is doing vs. some sell side analyst target. This could mean understanding who owns the stock and what matters to them. I've literally had sellside guys tell me that they get calls from HFs to discuss what other investors have asked them recently. I think that this can work, but then I think you really need to have a very good sense on who holds the positions, how they exit/enter and how they make their decisions.
Debitis ratione magni in cumque nemo qui et magnam. Magnam rerum ut ut. Tempore alias porro eaque vel.
Aspernatur officiis molestiae omnis eum est doloremque accusamus. Nostrum sunt architecto animi ut omnis nemo consequatur rerum.
Debitis velit perspiciatis iusto nemo ea et. Soluta deserunt molestiae quos. Molestiae nesciunt quis tenetur fugit expedita numquam doloribus. Ullam aperiam aliquam incidunt dolorem in non. Perspiciatis velit adipisci sequi occaecati.
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...
Qui voluptas ut ea laudantium doloremque culpa distinctio expedita. Soluta officia ut nesciunt provident nam tenetur. Architecto enim libero est consequatur consequuntur. Magnam optio blanditiis sed saepe.