How HFs Use Multiples
Very silly question but I’m new to investing and haven’t been able to find an answer anywhere else. I know using a DCF to derive a genuine PT is not common in HF and instead they slap a multiple on projected financials. Obviously how the financials are projected is highly thesis-dependent, but how exactly do you go about assigning a future multiple? Thanks!
This is the job - but generally valuation happens in context. So you look at historical ranges and narratives/metrics, relationship to peers and relevant pairs, alongside comparing the estimates you have. Art not science
Thanks so much. How do HFs approach IPO investments or less mature companies since they can’t refer to historical ranges? Is it generally how peers trade?
Never dealt with an IPO, but would imagine peers would be the closest bet. At an HF, you're just trying to make $$$, not get the multiple bang on. If its a beat and raise or a guidance cut, you can model the multiple directionally (vs historical range and relationship to peers). If you have a gut feeling on the multiple even after the quarter has traded, you can hold and let it compound.
Does DCF converge to PE multiples or do PE multiples converge to DCF?
Both. You can have models that spit out a PE based on your assumptions or have PE influence your PT.
Dude let me tell you something, I am an actual PM with more than 7 years experience in L/S equity at a pod. Nobody cares about multiples unless they are at extremes. If you're trying to convince your PM to buy a stock because it trades at 18x and you think it should trade at 22x, you're doing it wrong. Unless it's so cheap or so expensive to be optically unacceptable, arguing about nuances in multiples doesn't make you money at all. Stocks are much influenced by macro these days so individual company specific stuff matters less in day to day volatility.
the more relevant application is just using today’s multiple to price your non-consensus earnings/sales estimates?
What do you care about then, top3?
Just take today's multiple. E.g. "AMZN is trading on 4x FY2027 sales of $400bn". If your number if $450bn by 2027, multiply it by 4x. The exercise is to illustrate how your variance in the estimate drives your expectation of an upside or downside.
Theoretical underpinnings is the gordon growth model (basic DDM). EV/FCF = 1/r-g. If r = 8% and g=5%, EV/FCF = 33x. In real life, what matters is that you understand the relationship. In theory, you could look at the starting sales estimate of a stock in their s1 and find a strong r2 between subsequent sales estimates and the multiple.
Ultimately, the most useful insight is that the market multiple expands when the average investors estimates are higher than consensus and vice versa. Use that wisely. :)
Do you guys always bake in multiple expansion when you have upside on earnings? I tend to get my target range by having 50% of my upside from estimates and 50% from valuation
The thinking is that if the company beat, the stock will probably rerate but I’m wondering if this is wrong to do that.
A stock isn’t going to rerate just because it beat in the quarter. The beat must signal a better out-year dynamic to drive a rerating. Often if a co dropped the ball in q1-2, it’ll go back up to a higher/lower multiple from q4/(prior qtr) if it signals its back on track.
As an analyst it’s part of the job to figure out which year is being hit, and how the multiple would change on consensus numbers. E.g. going into TSLA earnings you know Elon is going to talk up (1) robotaxi rollout, (2) new market in India, (3) year end model announcement, etc. Well, (1) won’t do anything until it scales — so fy 26-27+ at earliest, while (2) India could probably be accretive in fy26 (though w/o new giga factory, production remains constrained). For (3), would be a fy27 story earliest (cybercab not expected to ramp until end of fy26 and that was announced last year).
So now I have my “bingo” card of what to expect on earnings. Next step is to create a reasonable range of estimates for each critical factor (1, 2, 3), and then roll them up into a set of scenarios (bull/base/bear/ad hoc). Should then price the stock on each scenario, and then see what your views imply about estimates — which vs. consensus tells you where you think the multiple will go (up if you think estimates higher and vice versa).
Deserunt dolore ad qui quibusdam. Ex ab qui veritatis vel. Omnis esse fugit sed corporis impedit esse. Velit placeat nulla voluptate harum tenetur.
Ipsam maxime illum voluptatum. Iste aut sit deleniti quas aut autem praesentium.
Amet quasi aut possimus ut. Perferendis nostrum eveniet quae quisquam. Doloremque blanditiis ut ipsa nobis quidem fugit. Pariatur veniam non odit natus eos.
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...