Fair enough. The times I need to model a company, that’s more or less how I deal with it too. I just wonder if there’s something better that you guys with more company-modelling experience do.
When doing top-down stuff, extrapolation and mean reversion has similar problems, but since profits and growth tend to get redistributed among companies in a sector over time, you tend to get an averaging effect that works in your favor from a macro perspective. It’s still unsatisfying, but not to the same extreme.
I think hedge funds (at least on the equity side, where I’m familiar) are a great investment vehicle, especially now when there’s been such a prolonged flight from risk, fixed income remains such a crowded market, yields are at an all-time low, and ultimately I think people are paying too much for current income/dividends rather than longer-term value. That’s certainly created some buying opportunities in my view.
Let me answer your question from the standpoint of someone that’s actually helping to market our hedge fund right now. Of course, I’d rather be picking stocks all day and night, but we need to raise funds and ultimately all hedge funds have to gather assets in order to scale. When I think about what we need to persuade LP’s on, it’s more than just our performance – it’s about making our investment process and philosophy relevant to them, proving our track record, and establishing that we’re people that they can trust. If you were an LP, you should want to think about the same things but from the other side of the table. In addition, if you’re driven more by monthly/quarterly numbers but a potential hedge fund really believed in deep value, then you have some real decisions to make about whether you can also be disciplined and trusting enough to put your money with that fund manager, or whether his style will be at odds with what you’re trying to accomplish.
At the end of the day, I do think the right hedge fund can be a great investment especially with fees on most funds coming down (rather than going up, given the challenging market). However, as with any good investments you have to be thoughtful about who you’re parking your money with. I realize this isn’t a “yes/no” response to your question but your question is a good one and definitely not one where a simple answer would be sufficient.
Depends on the industry. Some companies will have more transparent and frequent metrics than others (i.e. backlog, same-stores numbers, IMS data, etc.) that you can track. You can also get more clarity around things based on channel checks and customer surveys. Sometimes it does feel like you’re putting a finger up in the air; other than that it’s still about making educated bets. But one example is, on the topic of pharma, let’s say that a company’s drug sales have been tracking about +4% Y/Y per month. Expecting that drug sales should somehow tick up to 10% Y/Y at some point would be in most cases a reckless assumption. Sure it happens and if you’re going to make that assumption in your model, you should be speaking with docs, sales reps, etc. in the interim to get some real evidence supporting your bullish hypothesis in front of the IMS numbers.
So yeah, assume conservative growth if you care about margin of safety, but frankly what we try to do is we care less about conservative numbers – we care about accurate numbers. And I’m ready to do the research to figure out what that accurate number is. And if I feel like I can’t pin the growth down, then I have to defer to understanding the intrinsic value of the assets better because investing too much in growth without having a solid reasoning to back it up is essentially speculating and not investing.
How do you develop accurate numbers for FCF projections 10 years, 20 years, 30 years, and 40 years from now based on backlog/same stores numbers/CMS data?
I almost never model anything and I find modeling to largely be a waste of time. I do track out segments via a simple spreadsheet and will sometimes do a basic sensitivity or scenario analysis to understand leverage and the possible range of outcomes, but I don’t create earnings or DCF models. Most of my projections are back of the envelope where I say, XYZ should earn X based on A, B and C factors, and here’s why I’m confident in those factors.
The stock market is a discounting mechanism for the next 6-12 months. What the market does really well is incorporate all “obvious” information such as guidance, consensus, recent trends, news items, and so on. A particular stock will then often trade between X and Y based on a range of likely outcomes (e.g., people think the stock will trade between 10 and 12x EPS and should earn somewhere around $1.00 to $1.10 this year, so the range of the stock is between $10.00 and ~$13.00).
What most investors do is invest at $10.00 because the stock is “so cheap” as it is trading at the low end of its historic (or recent) multiple range. They then hope the stock’s multiple expands or the earnings grow and move them up to fair value. But the truth is that the market is not a perfect pricing mechanism, and any asset has a range of value – the value of an asset might be pretty close to perfectly knowable if you had inside information, but in the absence of that, you can only assign a range (the idea of an exact price target is sort of laughable). The range is greater or lesser depending on the business quality, stability of sales and earnings, and so on, but it’s always a range.
The problem with “range investing” is that you don’t really know what’s going to happen in the short run – the range represents uncertainty, and frankly, playing the range game is very competitive anyway since that’s what most people are focused on. I automatically assume that I’m not smarter than everyone else focused on short-term results. If the company rolls over and craps the bed, then all bets are off at 10x $1.00 of EPS and your thesis is rubbish – and if I only own the stock because I believe it should trade at 12x, now I have to sell at a loss. And since most companies 1) don’t grow value that well in the immediate-term anyway, and 2) the market prices the range well for most stocks, you really are not getting a bargain here – the risk / reward is usually pretty balanced.
What I focus on is predicting capitalistic outcomes, focusing on dramatic changes in a business. I’m looking for really material developments over the next 2-3 years where I can reach a good understanding that the stock’s current range is completely wrong such that the debate is no longer about 10x or 12x – it should be 20x, and by the way, the earnings are going to be $2.00. So it’s really about understanding the underlying business and how it creates or destroys value over time, and not so much about “the stock” per se.
If I can find that wide delta, I don’t have to be perfect on the timing, because the fund I work for has permanent capital and we hold a wide basket of these stocks. If I can nail the timing, that’s a bonus.
But anyway, I think some of these outcomes are knowable with a high confidence level, and I believe that the market does a poor job overall of focusing on these sorts of changes. The stock market has no imagination – it’s great at pricing in what has already happened (often instantly) but it’s bad spotting huge changes before they occur.
Normally 12-18 months for longs, though I do grow/trim a position around earnings and other key events and depending on how my perception of risk/reward profile changes. For shorts, closer to 6-12 months because given the upward bias of the market and most investors, it’s important to really understand upside risks and how that impacts valuation especially on the short side.
I guess in some way I think of myself as a “long-term value-oriented investor, one quarter at a time.”
bchad - For example NSR is such a stock I have no idea how to predict future cash flows. So I just plug in a conservative growth rate of 5% declining to 1% over the next 50 years…and it still comes out undervalued. In essence, by lowballing I get a margin of safety, and if it grows faster (as it should), in theory I should make a great return.
Same for GOOG. I lowballed a growth rate of 6% declining to 3%…and came out undervalued compared to (then price) of 575. (Google is a great stock for a beginner to analyze btw).
I don’t know if these will work out like I’m hoping…but it’s one possible way of projecting future growth.