The Paradox and the Lifeboat

For a chronological index of my path to the CFA, click here.

In her final hours, when her fever was high…
She reminded us of having earned her CFA designation…
I think it brought her ….comfort.

The Furball pg. 729

It is said that a really good Zen master is one who is able to destroy every concept you believe in, leaving you ultimately with nothing to cling to, so that you’re then ready to approach the investigation into the nature of reality with a fresh, beginner’s mind.

Well the CFA program is continuing to be my Zen master when it comes to investing. To quote the Volume II study guide, …which is quoting Buffet, …who is quoting Darwin:

Ignorance more frequently begets confidence than does knowledge.

Or to turn it around, the more you learn the less you realize you know.

I’ve already written about it before but it’s back on my mind today after finishing Volume II of the Level II study guides. It’s what I call the Great Paradox of the CFA program – that instead of each successive study guide chapter resulting in a new technique or refinement to add to my stock-picking toolbox, I’m instead mostly just throwing away the few tools I did have and not replacing them with much of anything… yet.

Luckily I did not drown as my Toolbox Titanic hit the Iceberg of Knowledge but rather escaped to a lifeboat that’s been fine for the time being. I’ve even been comfortable here in the midst of the Force 12 winds that are this current bear market! Her name is HMS Asset Allocation. Yes – indexing.

To paraphrase some excerpts from Morningstar, …who was interviewing Jack Bogle, …who was sort of quoting Winston Churchill, …who was critiquing democracy:

“It (indexing) is really not an intelligent way to invest, which leaves us with the question of why, year after year, does indexing beat 75% of all the active managers?

Indexing is a terrible way to invest, but it’s a better way than any other way that has ever been devised.”

Link to the brief video interview here: here.

So I’m sure some will get a good laugh at this fancy-schmansy Level II CFA candidate now investing in a very Level Zero CFP-ish fashion – diversifying broadly, re-balancing periodically. But to again quote Buffett…

Wide diversification is only required when investors do not understand what they are doing.

So is that what I’m effectively saying by my actions – that I admit to not knowing what I’m doing?

Not necessarily. On one hand I have managed to beat the S&P 500 with slightly lower risk (monthly standard deviation) over the last few years, including even this terrible year. But I can’t say I’ve done the quantitative due diligence to see if my results are statistically significant given my sample size.

And each passing CFA study chapter, possibly by focusing on worst-case scenarios, seems to point to fairly large holes in my previous approach’s defense against selecting portfolio “torpedos”, as they call them. They’ve been rare, but I have indeed been hit by a few (MTEX and FMD).

So why the flight to indexing? And why not step out of the lifeboat and just replace my ignorance-was-bliss toolset with a new, improved CFA version?

Herein lies the problem. The old tools were easy to understand, develop, and use. I’d even automated all of it by writing a computer script that retrieved all the bits of information from various websites that I used in the valuation and decision-making process and computed current versus intrinsic value. Give me a ticker and I could give you a buy / hold / sell rating along with a margin of safety estimate in about 5 seconds.

And the new CFA-style approach is just not that simple. That’s because all the financial ratios, historical & predicted earnings, growth rates, etc. cannot simply be derived from the companies’ financial statements as is. Your figures have to come from your own set of “modified” financial statements, which is the result of reading what is released by the companies like a sleuth and making various adjustments here and there.

Did I say sleuth? I meant to say skeptical sleuth. A grumpy, skeptical detective with a “guilty until proven innocent” mindset. You’ve gotta go through all footnotes and MD&A with a magnifying glass – compensating for off balance sheet debt, extraordinary income that has been counted as ordinary, ordinary expenses that have been counted as extraordinary, etc., etc., etc.

Yes Level I’s Financial Statement Analysis, with all its operating vs. capital leases and expensing vs. capitalizing, was just the appetizer. Level II continues the fun with special purpose entities (think Enron), effect of foreign subsidiaries and currency translation, and on and on.

The Volume II study guide makes the point again & again. When it comes to financial statements released by companies, every number except cash is the result of choice and is arrived at through the discretionary accounting policies and decisions of the management. Does it look like your company is going to have trouble meeting the street’s estimate of quarterly earnings? No problem, cut back on the amount you allocate for returned merchandise and extend the estimated useful life of some assets in order to reduce depreciation expense. Voilá- earnings now match what the street was expecting!

As I once read (source forgotten), earnings cannot be taken at face value because they’re the last item that appears on the income statement – a function of all the line items above it which may either be manipulated, smoothed, or based upon best guesses and estimates of managers. In fact Volume II says that that financial statements prepared under GAAP are “riddled with estimates”. Oh joy.

So due-diligence fundamental analysis is going to be messy, hard, and require getting my hands dirty. I can’t imagine being able to properly analyze more than 1 company per day. I’ve already tried to do a few businesses on my own but got quickly bogged down. Like opening up a German newspaper with much confidence after taking German 101 and 102, only to quickly be overwhelmed by how far you have left to go until fluency.

So given my engineering background one might wonder why I don’t escape to the quantitative dimension – building computer models from historical data and/or unadjusted ratios – skipping all that financial statement deconstruction and reconstruction.

Unfortunately the answer here appears to be the same as for fundamental analysis. To do it properly is not easy. Sure I’m a pro at coming up with a model to fit a particular data set, but it’s quite another step to do the due-diligence to see if the model actually does have measurable predicting ability, and you’re not getting fooled by one or more of many pitfalls outlined in the CFA study guides (serial correlation, multicollinearity, heteroskedasticity). Never heard of those last two? Neither has my spell checker!

So in quant world the bottom line is, it will take months to develop models that I can have a high degree of confidence in, and they will need to be continually updated and re-tested for accuracy and stability. The data itself will need to be periodically monitored to insure no wild changes in mean, variance, distribution, etc.

Do you get what I’m saying here? To do stock analysis properly is going to be hard, hard, hard! Sure I’m getting exposed to all the right tools in the CFA course but I have to internalize and understand how to use them. Reading a German-English dictionary from cover to cover doesn’t begin to make me fluent! 🙁

I think I see how this is going to play out. Gradually I’ll master enough of this to start doing some good analysis of small, simple-to-understand US companies, perhaps taking small positions in a few. As I gain experience I’ll slowly branch out to larger multinationals, taking the portfolio increasingly away from the cores (ETFs of various asset classes) and more into satellites (individual stocks). It’ll be a long road but I do really love this stuff.

Time for another quote:

Except by coincidence, book value equals neither the market value of the firm nor the fair value of its net assets.
– CFA study guide, Level Two, Volume Two pg. 295

There is a common workaround I often have to use in probability theory. Sometimes calculating the probability p that a particular event will happen is incredibly difficult, but computing the probability that it won’t happen p’ is easy. So I just compute the latter and recall that 1-p’ = p. Problem solved. 😉

Insight into a similar workaround that I think I’ll be using in financial statement analysis was provided to me by David Swensen’s excellent Unconventional Success (a book on indexing). He talks about the difficulty analysts face in looking at a company’s assets and figuring out how much they’re really worth. How much are the global facilities owned by Ford worth? What about its famous (intangible) trademark?

Yes even Yale’s expert endowment manager says that the most diligent analyst would recoil at the thought of trying to conduct an asset-by-asset valuation inventory of everything Ford. But what if you skip assets and move to liabilities…?

The market value of a company’s debt and equity is usually much easier to determine. And by recalling that Assets-Liabilities=Shareholder’s Equity, it’s 1-p’=p all over again!

Ah well, time to crack open Volume III. Happy studying everyone. And here’s the promised running tally of time spent preparing:

Read/Work Problems of Volume 1: 26 hours
Read/Work Problems of Volume 2: 15 hours
Total Preparation Time So Far: 41 hours

Disclosure: the author currently, and regrettably, owns a position in FMD.