I’ve read that price multiples are misleading unless you know the absolute valuation of its benchmark. Now when I think about doing DCF for the absolute valuation, I realized analysts may have the same assumptions except the forecasting period which may result to different intrinsic values.
How would you guys(equity analyst) remedy this? Did I misunderstand something?
Assumptions only apply to the forecast period by definition (the rest is known). The whole point of investing comes down recognizing the primary drivers and risks in a business and making better assumptions than the market (which leads to better forecasts).
Personally, I don’t spend very much time at all on models, in fact I think they generally do more harm than good. Multiples can be helpful as a barometer for prices that are too high, but generally I focus on isolating risks and drivers and forming my own views on them. Sometimes it’s as simple as recognizing that the drivers are likely to be more volatile than the market’s unimaginative expectations (pretty common) and understanding when this may be very detrimental or positive for a particular firm (i.e. recognizing imbeded optionality).
Thank you for replying to my post.
Please correct me if I misunderstood, but it sounds like you prefer the qualitative analysis more than the valuation. I thought qualitative analysis was is to understanding the business ultimately for a more complete valuation.
Actually my question is due to an analyst using a forecasting period of 20 to 25 years. I was wondering how valid or accurate that was. Isn’t the period unconventional? (Maybe the analyst is doing this to increase the stock’s value and not exactly based on logic but an arbitrary number) What’s your opinion on this?
DCF models often get split into a near term period of say 5 years, and the longer term. Sometimes you see the longer term section split into a 20-30 year forecast and a terminal value. So the analyst inputs their assumptions for the next 3-5 years which they will be reasonably confident in and then makes broad assumptions for the future period such as sales growth being equal to nominal GDP growth, constant margin etc.
The ultimate ‘value’ that the DCF model generates will vary enormously depending on the assumptions you make in your inputs. There is no right answer. Where models are useful I think is to give you a sense of how the current share price compares to an intrinsic value based on what you consider to be reasonable assumptions. The model will also allow you to conduct sensitivity analysis to see how the value changes when different inputs vary.
Almost never do analysts model out beyond 5 years. Increasing the period of forecast will neither increase the accuracy nor the actual value the DCF model produces. Anything further than 5 years is essentially just fooling yourself about the accuracy of what you’re doing, unless there’s a known regime altering event occuring beyond that time period, which is rare. It’s the classic garbage in, garbage out. As an example, Warren Buffett and Charlie Munger have never used models to invest.
Qualitative analysis is where most quality buy side analysts spend the majority of the time. It is also in most cases the most difficult aspect to master for difficult investments. Any monkey can model a DCF, it’s a basic skill. Some buy side analysts don’t even model as part of their process as a result. The model is unnecessary in many cases unless you’re doing PE, leveraged, or capital structure deals. It’s also often important for modeling liquidity and determening sensitivities. As an analyst you know what the inputs and assumptions are that the rest of the industry is using and you know everyone can build a model. So for me I spend most of the time figuring out where we may disagree and the rest follows somewhat intuitively from there.