How long to stretch a DCF?
Prospective Monkey here. At my college, my professor always talks about a 15 year DCF, which I find ridiculously long. There is barely any visibility 2 years out let alone 15.
I was wondering how do people actually working at HFs frame their DCF assumptions and timeframe? And how valuable is it really when coming to a final conclusion about the stock?
Easy - don't use DCF
so can you provide a framework or process what you would use in place of DCF?
FCF yield only thing I use
Five years historical, one year projected, four years forecast.
What is the difference between your projected vs forecast? Is the 1 yr projected from current consensus estimates and the forecast is your own forecast assumptions?
DCF is very helpful for u to understand valuation of any company; however, it is not that useful when dealing with recent years. So try this as a thinking pattern. For example, Car industry company has a different DCF VS consumption company.
Your DCF should go out far enough for the company to reach its steady-state level of growth and profitability to be done correctly, there is no one-size fits all number of years. Additionally, pertaining to your comment on visibility, the bulk of the value of a company lies in its ability to continue generating cash flows far out into the future. Just because that is difficult to predict does not make this any less true, and it is implicit in market-based statistical value measures.
Theoretically however long it would take the underlying business to reach the so called steady state.
Also make sure you model out a full business cycle so that your terminal doesn’t end up at a peak or trough.
Take a read of Morningstar’s DCF methodology - they’ve taken a ROIC and cost of capital convergence approach, and broke down the “path” to three stages. Quite methodical and interesting but could be overly complicated if you are just doing a quick analysis.
http://csinvesting.org/wp-content/uploads/2015/02/CommonErrors-in-DCF-M…
I have this printed out and on my desk.....
Depends on the industry. The implicit assumption with a DCF is that the run-rate business will exist in perpetuity. That is not the case in some industries (mining, energy, pharma) as assets are finite, especially if there is a relatively limited number of them in their portfolio. In these cases you should in theory model out to the end of the asset life, especially as (given time value of money) the tendency is to extract as much cash flow as you can in the initial years, make later ones much less profitability.
Tl;dr - steady state is not a concept that applies to all industries
Depends on the industry. How long can you assume a steady state for the business? Usually around 5 is more realistic, but if this is a pre-revenue biotech or something like that, it will go all the way to the patent cliff. Lots of massaging the numbers!
My suggestion: skip DCF when possible. Good luck finding the "steady state" of a business.
I like DCFs conceptually and there are ways to use them for certain things, but I wouldn't put a lot of faith in the value you come up with.
Forecast out FCF/share or whatever is relevant for the business you're working on for whatever time frame is relevant to your thesis, put a multiple on it (you need to understand what drives a multiple, don't just pick it out of thin air) and IRR it. At least bull, base, bear cases. Is there asymmetry to the profile? Is it legit? Is it attractive?
Not a HF guy, but I assume the same principles would apply as in IB and consulting - stretch it out as long as makes sense. Nobody builds 20-year projections for seed stage startups. But also nobody builds 3- or 5-year projections for power stations and toll roads.
The logic of a DCF is that the business is not yet at steady state. It's still growing at an un-natural rate. So you want to know how much cash you'll get during the stub period, and then you can apply a terminal multiple which is appropriate for a business that's going to grow at a stable rate thereafter.
Thus you do a long enough DCF to get to steady state.
It's true that you don't have much visibility, but that's just as big of a problem with a shorter DCF. Say you expect a business to take 15 years to get to steady state, but then you only do a 5 year DCF because that's all the visibility you have; well now your terminal multiple is messy because it includes 10 years of un-natural growth along with the steady-state period after.
I don't do DCFs hardly ever, but if I'm going to do one, I'm going to project the whole period of unnatural growth. 15 yrs if that's what I expect.
D C F?
D C F?
D C F?
Some really great information here from all the replies.
i use an approach very similar to what Morningstar uses. Three stages, with terminal assumptions relating to moats. But this is one tool in a big tool box and that’s the message I would extract from what everyone posted.
A long DCF is great for project finance models. For company analysis, part of the tool box. It can help with running upside and downside scenarios for extreme bookends. But you will want other tools to determine valuation and entry points. Those models can be very sensitive to the long term discount rate as well. So it can really skew the value of growth companies.
I pretty much only do DCFs but that’s a function of investing in businesses that are not in any kind of steady state.
Also, saying something trades at a 10% FCF yield and will magically trade to 5% is a stupid way to invest.
lastly what a lot of people don’t seem to realize is that at higher discount rates the PV impact of going out to 15, 20, 30 years is meaningless. You capture almost all the value in 5-10
Iusto maxime id minima. Ab quisquam itaque necessitatibus. Velit incidunt vero totam vel sequi necessitatibus sint voluptatem.
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...