i came across a business insider detailing socgen’s research which expects a 15% drop in the s&p within the next year, due to a movement in more restrictive monetary policy by the fed in q1. Part of this thesis is predicated on similar market moves when qe1 & qe2 were withdrawn. What I have issue with is the second part of the research which indicates the swiss and american equity markets are currently the most overvalued based on a p/bv against ROE regression. Intuitively it makes sense that the s&p is overvalued as we have seen consistent increases in the index on relatively weak earnings additionally we havent seen a significant improvement in american economic strength as indicated by the capacity utilization index, unemployment and pmi.
that being said my discomfort is the use of an accounting based valuation metric like p/bv because
1 accounting standards vary amongst jurisdictions so they aren’t 100% comparable
2 book value is based largely on historical cost rather then fair market values
3 accounting standards allow for alot of judgements being made, judgements which increase the potential for manipulation.
so my question is, if you are a international value investor what type of valuation metrics/practices do you employ to get around these limitations?
My favorite metric is EBITDARDSGACOGS, otherwise known as sales. Revenue is hard to manipulate. It’s not impossible because somewhere some company is figuring out how to create fake shipping orders to troll their auditors and the SEC (this works btw, most auditors are terrible and the SEC is woefully incompetent) but in general, sales is the least fudgey out of any financial metric.
Book value is subject to historical cost and depreciation assumptions.
Any of the numbers below sales on the P&L can be manipulated based on accruals and timing issues. Of course everyone is aware of earnings manipulation as a concept but there are some sneaky games that can be played with any of the line items that ultimately lead to earnings.
The cash flow statement can be manipulated based on timing of capex and balance sheet items.
However, sales are the least likely to be manipulated and sales are sales, there isn’t much difference between international accounting rules on sales unless a specific company has wonky revenue recognition practices.
So P / Sales and EV / Sales are good way to go. However, I have to be “that guy” and point out here that market cap / stock price is also fairly easy to manipulate. Companies actively try to massage their stock prices, the global settlement is dead so that underwriter-research analysts can flagrantly and openly defend stocks of their clients, and paid touts will often blast out manipulate reports on SeekingAlpha and other sources.
If I’m the CEO of a public company, I can pay a research distributor somewhere around $300-600K a year depending on the specifics to help me get my story out. I can then tell the Street, “OMGWTFBBQLOL!!! LOOK AT TEH GUIDANCE!!!1” even if the guidance doesn’t make any sense whatsoever. If a company comes out with stupidly high guidance, the market will focus on “high” instead of “stupidly” and the stock price will go up. Eventually this has to lead to a stock price reversal, but usually it gives the executives time to cash out some stock and/or use the stock as currency for a (likely dumb) acquisition.
The bottom line is it’s a dirty game and most companies are manipulating at least some of their metrics most of the time to a greater or lesser degree. The key is to avoid the really horrendous stuff.
I’ve always shied away from traditional fundamental analysis, because it always struck me that it’s an enormous amount of work to make questionable judgements based on highly manipulated and unreliable data. Unfortunately, there are just many more jobs doing it, so maybe that was a bad choice for me, but it’s hard to do a job well when you think it doesn’t really add a lot of value.
i do have to say that I like bromion’s general approach in posts he’s placed here because he is looking for strategy and organizational behavior which is at big odds with what anying the statements could possibly justify. These are the kinds of things that a percent difference here or there on some number isn’t going to make much of a difference on, and things that a zillion computers calculating statistics aren’t likely to identify, and that highly paid monkeys blindly following whatver the company guidance says are going to miss.
It seems to me that investment opportunities are going to break down into those things that you can calculate with automated quantitative processes and those things that you need human understanding to figure out, and bromion is being intelligent about where the opportunities are on the human-thinking side.
i do thing P/CF is one of the better metrics, but companies can still figure out ways to manipulate the breakdown of CF. P/S does seem less manipulable than other stuff and Hussman says it works pretty well as a macro aggregate indicator too. It presumably fails when companies are redesigning their processes, because it can miss out on internal efficiency improvements. Probably works well for stable and mature industries.
I feel uneasy looking at P/S instead of EV/S because of differences in capital structure. When peers have a capital structure that is close enough, I guess it doesn’t matter that much. The same can be said when comparing a company against its own past. But, otherwise, I think P/S tends to be less reliable. Am I missing anything?
I agree, I prefer EV / Sales but it can still have some difficulties with companies that have unusual capital structures. It is harder to apply to highly net indebted companies and heavily net cash rich companies. I actually do not like companies with a lot of net cash most of the time anyway. I use all the metrics though, it just depends on the situation. The one I use least often is EBITDA since that is the metric that Wall Street uses to justify overpaying.
I like EBITDA because I think it shows trends better than FCF (which is lumpy), but I don’t think it is very useful for valuation. I focus on FCF for valuation, but since it can be erratic, I find the historical ratio of FCF/EBITDA and multiply to latest FY EBITDA to come up with a “base rate”…yeah very scientific I know.
Also, I believe for technology firms, you should consider acquisitions to be CapX. Also I believe in backing out share repurchases from FCFE if you’re looking to add per share growth in metrics to your model.