The following was published on 6/1/2004 at RealMoney.com

=-=-=-=-=-=-=-=-=-=-=-=–=-=-=-=-=-=-=-=-=–==-=-=-=-=-=-=-=-=-=-=-=-=-=-

Investing Strategies

 

Models that work well on data about the past may not work in the future.

Check methods for weak points, like overfitting or ignoring illiquidity or business relationships.

Keep in mind some practical considerations when testing a theory.

Other Areas of Data-Mining

In 1992-1993, there were a number of bright investors who had “picked the lock” of the residential mortgage-backed securities market. Many of them had estimated complex multifactor relationships that allowed them to estimate the likely amount of mortgage prepayment within mortgage pools.

Armed with that knowledge, they bought some of the riskiest securities backed by portions of the cash flows from the pools. They probably estimated the past relationships properly, but the models failed when no-cost prepayment became common, and failed again when the Federal Reserve raised rates aggressively in 1994. The failures were astounding: David Askin’s hedge funds, Orange County, the funds at Piper Jaffray that Worth Bruntjen managed, some small life insurers, etc. If that wasn’t enough, there were many major financial institutions that dropped billions on this trade without failing.

What’s the lesson? Models that worked well in the past might not work so well in the future, particularly at high degrees of leverage. Small deviations from what made the relationship work in the past can be amplified by leverage into huge disasters.

I recommend Victor Niederhoffer and Laurel Kenner’s book, Practical Speculation, because the first half of the book is very good at debunking data-mining. But it also mines data on occasion. In Chapter 9, for example, the authors test methods to improve on buying and holding the index over long periods by adjusting position sizes based off of the results of prior years. Enough results were tested that it was likely that one of them might show something that would have worked in the past. My guess is that the significant results there are a statistical fluke and may not work in the future. The results did not work in the recent 2000-2002 downturn.

As an aside, one of the reasons Niederhoffer’s hedge fund blew up is that he placed too much trust in the idea that the data could tell him what events could not happen. The market has a funny way of doing what everyone “knows” it can’t, particularly when a majority of market participants rely on an event not happening. In this case, Niederhoffer knew that when U.S. banks fall by 90% in price and survive, typically they are a good value. Applying that same insight to banks in Thailand demanded too much of the data, and was fatal to his funds.

What to Watch Out for

Investors who are aware of data-mining and its dangers can spot trouble when they review quantitative analyses by looking for these seven signals:

1. Small changes in method lead to big changes in results. In these cases, the method has likely been too highly optimized. It may have achieved good results in the past through overfitting the model, which would interpret some of the noise of the past as a signal to return to the earlier analogy.

2. Good modeling takes into account the illiquidity of certain sectors of the market. Any method that comes out with a result that indicates you should invest a large percentage of money in a small asset class or small stock should be questioned. Illiquid or esoteric assets should be modeled with a liquidity penalty for investment. They can’t be traded, except at a high cost.

3. Be careful of models that force frequent trading, particularly if they ignore commission costs, bid/ask spreads, and, if you are large enough relative to the market, market impact costs. These factors make up a large portion of what is called implementation shortfall. In general, implementation shortfall often eats up half of the excess returns predicted by back-testing, even when back-testing is done with an eye to avoiding data-mining.

For a full description on the pitfalls of implementation shortfall, read Investing by the Numbers, by Jarrod X. Wilcox.  Chapter 10 discusses this issue in detail. This is the best single book I know of on quantitative methods in investing.

4. Be careful when a method uses a huge number of screens in order to come down to a tiny number of stocks and then, with little or no further analysis, says these are the ones to buy or sell. Though the method may have worked very well in the past, accounting data are, by their very nature, approximate and manipulable; they require further processing in order to be useful. Screening only winnows down the universe of stocks to a number small enough for security analysis to begin. It can never be a substitute for security analysis.

5. Avoid using quantitative methods that lack a rational business explanation. Effective quantitative methods usually come from processes that mimic the actions of intelligent businessmen. Never confuse correlation with causation. Sometimes two economic variables with little obvious financial relationship to each other will show a statistically significant relationship in the past. Two financials merely being correlated in the past does not mean that they will be so in the future. This is particularly true when there is no business reason that relates them.

6. Look for the use of a control. A control is a portion of the data series not used to estimate the relationship. It’s left to the side to test the relationship after the “best” model is chosen. Often, the control will indicate that the “best” method isn’t all that good. And beware of methods that use the control data multiple times in order to test the best methods. That defeats the purpose of a control by data-mining the control sample.

7. One of the trends in accounting is to make increasingly detailed rules in an attempt (wrongheaded) to fit each individual company more precisely. The problem with that is it makes many ratios difficult to compare across companies and industries without extra massaging to make the data comparable. This makes thinning out a stock universe via screening to be less useful as a tool. For quantitative analysis to succeed, the data need to represent the same thing across different firms.

Practical Recommendations

There are many pitfalls in quantitative analysis. But three simple considerations will help protect investors from the dangers of data-mining.

1. Paper trade any new quantitative method that you consider using. Be sure to charge yourself reasonable commissions, and take into account the bid/ask spread. Take into account market impact costs if you are trading in a particularly illiquid market. Even after all this, remember that your real-world results often will underperform the model.

2. Think in terms of sustainable competitive advantage. What are you bringing to the process that is not easily replicable? How does the method allow you to use your business judgment? Is the method so commonly used that even if it is a good model, returns still might be meager? Even good methods can be overused.

3. If doing quantitative analysis, do it honestly and competently. Form your theory before looking at the data and then test your theory. Then, if the method is a good one, apply the results to your control. If you perform quantitative analysis this way, you will have fewer methods that seem to work, but the ones that pass this regimen should be more reliable.

The following was published at RealMoney on 5/28/2004:

-=-=–=-==-=-=-=-=–=-=-==-=-=-=-=-=-=-=-=-=-=-=–=-=-=-==-=-=-=–=-=–=-=-=-=-=-=-==-=-=-=-=-=-=-=-=-

Investing Strategies

 

Data-mining attempts to get data to give a sharp answer when one may not be present.

Technical analysis can involve data-mining.

Chance can make a method look better than it is.

Investors often get pitched quantitative methods for investing. These methods can be either fundamental or technical in nature and often have shown great results on a pro forma basis in the past, but when ordinary investors (and often, professional investors) try them out, they don’t work as well in practice. Why?

There are many reasons, but in my opinion, there’s one main reason: data-mining. I’ll define data-mining and give you practical ways to avoid it whether you apply quantitative methods or create new quantitative investment methods.

Data-Mining Defined

I never got my doctorate, but I did complete my field in econometrics in grad school. One of the things that they drilled into us was the danger of overinterpreting your data. As a mythical economist supposedly once said, “If I torture the data enough, I can make it confess to anything.”

When a quantitative analyst mines data, he repeatedly tests new hypotheses against the same data set. When the analyst finds an economically or statistically significant relationship, he stops testing alternative hypotheses. He may start to optimize the hypothesis that gave a significant result.

Data-mining, or as some call it, specification searching, attempts to get the data to give a sharp answer when no sharp answer may be present. Financial data are messy; there is a lot of noise and often not much signal. Every time data get analyzed, there is a small but significant probability that noise in the data will be interpreted as a signal.  Overinterpreting the data increases the odds that what the analyst thought was signal was actually noise.

Examples of Data-Mining

As examples, consider Michael O’Higgins’ Beating the Dow, which introduced and popularized his “Dogs of the Dow” theory, or James P. O’Shaughnessy’s What Works on Wall Street. In each of these books, different hypotheses were tweaked to find a method that would have produced the best result in the past.

The basic idea underlying the “Dogs of the Dow” theory has merit: Buy cheap, large-cap stocks. But in testing multiple theories, the cheapness metric was varied. Which is the best: low price-to-book, earnings, sales, cash flow, low price or dividend yield? Another factor that varied was which stocks would be picked. Would it be the top 10, top five, top one or even the second-best? How often would the strategy get rebalanced: annually, quarterly, or monthly?  With this many permutations, the strategy that ended up performing best likely did so accidentally.

What Works on Wall Street also contained some good core ideas (although it was a bit misnamed; it should have been titled, What Has Worked on Wall Street, but that would not have sold as well). Its core theory: Buy cheap stocks that have positive price and earnings momentum. But in this theory, the cheapness metric also varied, along with the methods for analyzing momentum — enough that more than 50 different theories got tested. The basic idea is sound, but again, the variation with the best result won only by accident.

… And Technical Analysis

Bloomberg has a back-testing technical analysis function [BTST]. It takes eight different technical analysis methods and shows how each would have performed in the past for a given security. Even if some of the methods had validity, if an analyst fed the BTST function a stream of random data instead of a real price series, the function would likely flag one of the methods as profitable.

Another area where I have seen abuse is in “services” that offer to identify “rolling stocks,” i.e., stocks that seem to oscillate between two predictable boundaries. This gives the potential for an investor to make quick and easy profits by buying at the low boundary and selling at the high boundary. The trouble here is that it is easy to identify stocks that have traveled in boundaries in the past, but the past is usually a poor predictor of the future. Results from following advice like this should be random at best, with the danger that your losses could increase if the conditions that created the temporary stability shift.

Data-Mining in Modern Portfolio Theory

Why do stocks always seem to do better than bonds in the long run? How much better should they be expected to do? These questions frame what is called the Equity Premium Puzzle. Academics who use data-mining assume that past is prologue and that initial valuation levels have no impact on the results for their forecast period. Back in 1999, I often commented that since 1926, we’d seen only one and a half full cycles of the equity markets. Naive estimates of the equity premium were popular among academics and practitioners then. We had not seen a second major bear market like that of the 1930s. The bear market of 2000-2002 has adjusted my view, but I am not convinced that valuation levels have returned to normal.

There are many societal and political factors that affect how much better stocks will do than bonds. People do not have infinite investment horizons; they will need at least some of the money at some point in their lives, so long-term total return averages are not indicative of what average investors are likely to achieve. Valuations matter, as do the current yields of bonds. Neglecting equity valuations and bond yields when doing asset allocation work will lead asset allocators to overweight stocks and bonds, which have done well historically but are unlikely to do as well over the next 10 years as the historic averages.

In a past job, I was a quantitative analyst for an asset manager that had a life insurance company as a client. There were a variety of derivative investments that got pitched to us that used diversification of different credit risks as a means for reducing risk. Often I would be shown a correlation matrix of past returns that showed high reductions in volatility from mixing different risky asset classes. I would ask the quantitative analysts on the sell side how stable the correlation matrix was, given how highly correlated most risky fixed-income asset classes were in 1998 during the Long Term Capital Management crisis, and afterward in the recovery. Most of the time, they hadn’t considered the question.

A Big Warning Sign

Anytime you see an analysis that relies on a correlation matrix of returns through some sort of mean-variance framework, be careful. My favorite target here tends to be a fund-of-funds, whether of the CTA, hedge, or mutual fund variety. There are several reasons for that.

First, there usually aren’t enough data to estimate the correlation matrix. Inexperienced practitioners do  so anyway, without realizing that they need at minimum, one data period for each unique correlation coefficient that they calculate. For example, for a correlation matrix of 10 return series, you would need at least 46 periods for the data, and really, you would want more than 70 to gain sufficient statistical credibility on a historical level.

Second, even if there are enough data to calculate correlation coefficients that are statistically credible, the financial processes that produce the correlation coefficients aren’t stable. Past correlation coefficients are poor predictors of future correlation.

Third, “past performance may not be indicative of future returns.” This is not only true of the level of returns, but also the variation of returns. It should not surprise anyone, then, that ratios of historical average return to the variability of return aren’t good predictors of the future ability of a manager to obtain returns with low variability of results. In short, Sharpe ratios (or reward-to-variability ratios) are, in my opinion, poor predictors of the ability of a manager or assets class to produce return and mitigate risk. Efficient frontier analyses draw pretty pictures, but they usually do not produce asset allocations that optimize the future risk/return tradeoff when the parameters are estimated from historical data.

Another data-mining villain is returns-based style analysis, which assumes that a manager’s true style can be discerned from the correlations of his returns with a variety of different asset class indices. Leaving aside the problems of multicollinearity and inability to develop confidence intervals on the constrained regression, the use of short historical data series might give a clear view of the past, but it is poor when used to predict how a manager will perform in the future. In short, the past correlations are poor for predicting future returns.

With academic financial research, it is good to remember that only the survivors get published, and surviving requires statistical or economic significance, either of which can occur for reasons of structure or chance. Data-mining allows marginal academics an opportunity to publish.

In the second part of this column, I will review some practical ways to assess quantitative methods and sidestep data-mining.

The following was published on 3/29/2004:

-==-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Investment Advice

Time horizon usually correlates with return size.

It’s good to have signposts as the investment plays out.

Free advice is seldom cheap.

In analyzing any advice, investors have to consider the adviser, personal character issues and the nature of the investment proposed.

In Part 1 of this three-part column, I focused on the adviser. In Part 2, I looked at issues centering on your personal character.

In Part 3 today, the emphasis shifts to the investment itself.

Many Things to Consider

Good investment recommendations give some idea of how much to play for and the likelihood of getting there, even if the appraisal of likelihood is subjective and squishy. Are we looking to scalp a dime, a buck, 10%, 100%, or are we looking to score the elusive ten-bagger?

Most often, the time horizon of an investment corresponds to the amount targeted to be earned. Under normal circumstances, gains are made a little at a time. Bigger gains ordinarily take more time. How long will it take to earn what is expected from the proposed investment?

What risks exist in realizing the value inherent in the investment? What could go wrong? Nothing is certain in investing, so beware of advice that tries to sell hard on the idea of safety. Appeals to safety, particularly with investments that are touted to earn an above-average return, are often dangerous. The price adjustments with supposedly safe investments that disappoint are sometimes severe. I experienced this firsthand with corporate bonds: The most dangerous bond was the one everyone knew was secure, and then accounting irregularities popped up. The price would drop 10% to 20%, and liquidity would drop to nil.

If the investment is going properly, what signposts will you see to validate that the investment idea is on track? Aside from price action, what will yield clues that the investment thesis is wrong or right? What should earnings look like? When is that new product going to be introduced?

What factors in the macroeconomic environment does the investment rely on? If inflation rises, what will happen? Does this investment resist recessions well? If the market falls, will this investment fall harder?

Finally, how well does this investment fit into your portfolio? Does it reduce risk for you, or increase it?  Too much of a good thing can be wonderful, but the more concentrated your bets become, the closer you must watch your positions. The higher the degree of concentration in a portfolio, the higher the amount of expertise relative to the market the portfolio manager must possess.

No one will give you all of this in advice, but these are things to keep in mind to aid in the evaluation of advice that comes your way. In general, a conservative and skeptical posture will serve you best. Keep a tight hand on your wallet, and remember that those who stay in the game the longest often do the best.

Finally, you can remember Ferengi Rule of Acquisition No. 59: “Free advice is seldom cheap.”

The following was published on 3/26/2004:

=–==-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Investment Advice

You have to understand the advice to use it.

Can you implement or monitor the idea?

Analyze your own personal motives.

In analyzing any advice, investors have to consider the adviser, personal character issues and the nature of the investment proposed.

Yesterday, in Part 1 of this three-part column, I focused on the adviser. Today, in Part 2, the emphasis shifts.

It’s About You

Do you understand the advice? There is no shame in not understanding every investment concept under the sun. Only rare individuals can do that. If you can’t understand what is being proposed, walk away from the idea until you can understand it. People who don’t understand an investment concept, but invest anyway, can’t react rationally to the volatility in the market, and they fall prey to fear and greed. They become the noise traders that professionals profit from.

Some strategies suffer from what I call “too smart for your own good risk.” In Britain, the phrase is “too clever by half.” This problem affects both individual and institutional investors. Some strategies are very complex, and some people are intrigued by complexity. I think most investing is simple, and complexity signifies a lack of understanding. The more complex a strategy is, the more likely it is to break down in one of its many steps. Be careful with complex strategies.

Can you implement and monitor the investment idea? Does it fit your character? I did risk arbitrage on an amateur basis for several years, but even though I did well at it, I found that the amount of time it took detracted from my family and work, so I stopped.

Some people don’t have the time, talent or personality for strategies that require rapid trading or rapid shifts in strategy. Other people don’t have the stomach for high-risk strategies, even if they understand how they work. You have to pick strategies you can sleep with.

Does the investment support your ethical standards? This applies to both the management and the business.  In general, your ability to make rational decisions in investing will be hindered if you are long a company that you think harms society. The same is true of management that you believe acts dishonestly, particularly toward shareholders. It doesn’t matter how cheap a company is: If you can’t trust the management, it will be almost impossible to unlock the value trapped there.

Also, from my personal experience, if management is dishonest to some other stakeholder group, such as customers, eventually shareholders will get bad returns. Dishonest management often has underlying business models that are unfavorable, and which they are trying to enhance unethically.

Analyze any personal motives you might have for making or not making an investment. I had a large number of usually intelligent friends who gave up their investment disciplines in late 1999 in order to buy into the bubble. Many seemed driven by envy of less capable friends who were racking up impressive profits on paper. Motives for investing that rest in uncritical admiration or dislike for another person and their prosperity usually lead to bad results.

How much of an unrealized loss could you take in the short run? Do you have the capability to carry the position through a rough period, even if the eventual result will be good? The answer depends on your liability structure. Do you need the value of the assets in question to throw off cash for you in the short run? Are you investing on margin, or have significant external debts to service? Safe is better than sorry here. At minimum, set stop orders if you can’t bear losses beyond a given threshold. It is better to avoid strategies that force you to take any action, so if you can’t take short-term losses, reduce the risk level.

Next time, in Part 3 of this three-part column, I’ll take a closer look at the nature of the investment itself.

I ran across this article, Value Investor or Value Pretender: Which Are You?, by John Mihaljevic who puts out The Manual of Ideas, along with Oliver Mihaljevic.  I appreciate what they do — you can learn a lot from their organization.

I told him that I was going to write this, and he said to me:

The piece was meant tongue-in-cheek but feel free to rip it apart

I will rip it apart, but gently, because every point he made is mostly true for value investors, but there are variations in the way that value investors operate, so you can do some of the things he says you can’t do, and still be a value investor — what matters is how you implement them.

There will be more parts to my “Education of a Risk Manager” series, and one of them will deal with all of the different managers that I met, and how much they varied in terms of what they thought were factors that mattered.

Thus, as I developed my own theories of value investing, I considered the range of opinion, and realized that there is a single model for value investing, but that it is complex enough that different parties use different approximations of the full model, and those approximations do better and worse in different environments.

Like a David Letterman-style Top 10 list, John Mihaljevic listed and described things that made you a value pretender.  Time to go through them:

Reason #10: You invest based on chart patterns

I don’t use chart patterns, but I do use momentum both positively & negatively.  There is decent evidence that investors are slow to react to new information, and so stocks with strong price momentum over 200 days tend to do better.  There is some evidence where there is lousy price momentum over a 4-year period, that things tend to mean-revert.

Granted, there is a tendency among some value investors to troll the 52-week low list.  I like doing that too, but you have to be careful, because maybe you are missing something that cleverer investors know.  The same would be true of short interest figures.  Whenever I see one of my stocks gain a high short interest ratio (shares sold short / volume, or % of mkt cap sold short), I do a review to see what I don’t know.  That’s why I am not afraid of the high level of shorting on Stancorp Financial.  This is a conservatively run firm that manages risk up front.  Even though disability claims rise when unemployment is high, they underwrite better than most of the industry.

There have been some very successful value plus momentum investors.  The balance is tricky, but blending two of the most powerful anomalies does bear fruit.

Reason #9: You assume multiple expansion in your investment theses

I never assume that, but if you are buying them “safe and cheap,” you often do get multiple expansion.  The challenge is figuring out where things are less bad then the implied opinion of the depressed valuation.

Reason #8: You try to figure out how a company will do vis-à-vis quarterly EPS estimates

I don’t do that either, but I have known some value managers that incorporate prior earnings surprise data, because past earnings surprises are correlated with future surprises.  Often, near the the turnaround point for a company’s stock, there are some earnings surprises.

Reason #7: You base your decisions on analyst recommendations

I have few arguments with this, except negatively.  Sell-side analysts are trailing indicators.  I like buying companies where the sell-side is negative, but not very negative.  With very negative opinion, there are often reasons to stay away, unless you possess specific knowledge that the sell-side analysts do not have.

Reason #6: You use P/E to Growth (PEG) as a key valuation metric

I’m sorry, but PEG works, if indeed you have the growth rate right, which is a challenge.  I do try to analyze sustainable competitive advantage for the firms that I own.  That often leads to growth.  Now I am a growth skeptic, so it takes a lot to make me pay up for growth, but occasionally I will do so, when the PEG is low enough.

Reason #5: You use EBITDA as a measure of cash flow

EBITDA is not cash flow from operations, or free cash flow, but it is a valuable figure in value investing when it divides into Enterprise Value (Value of Debt + Value of Stock – Cash).  Low ratios of Enterprise value divided by EBITDA are very effective at identifying promising investments — it indicates cheap assets, and in a time when M&A is hot, it can really pay off.

Reason #4: You would worry about your portfolio if the market closed for a year

I could live with the market closed, but there are advantages to having it open.  With any given stock, there are times in a year to increase or reduce exposure — if you have a firm idea of what the firm is worth, you can buy more during dips, and sell a little into strong rallies.  Short term (one month) stock price movements are fickle, and commonly reverse.

Reason #3: You make investment decisions based on the activity or tips of others

But Manual of Ideas tracks the 13F filings of great investors.  I get good ideas from the best investors also, but you have to do your own research.  Many bright investors chat with each other, and I had many occasions at the hedge fund that I worked for where I disagreed with a friend of the boss.  I was right more often than I was wrong.

Perhaps a better way to phrase it is “choose your idea generators wisely, but do your own research as well.”

Reason #2: Your investment process centers on the market opportunity

This is largely true, but when I know a industry or sector is in horrible shape, I often buy the strongest name in the industry, realizing that they will do well as the competition dies, and they don’t.  Also, there are times when few recognize that pricing power has shifted, and it is time to take a position on a misunderstood industry that is about to grow faster than expected.  Particularly with cyclical companies this idea can be promising.

The same applies to countries where the markets are washed out.  Don’t try to time the bottom, but when a country is cheap, buy a promising/safe company in the country after things have turned up for 100 days or so.

Reason #1: Your investment theses do not reference the stock price

At some points, I like to own companies with strong management teams relative to their industry.  I will let valuation stretch at those points, because there is more of a sustainable competitive advantage there.  You get more positive surprises, and that definitely aids total returns.

That said, a focus valuation is key to all investing.  The only thing more important is margin of safety.

Margin of Safety

There are three elements to margin of safety:

1. Sustainable Competitive Advantage (Strong Gross Margins)
2. Strong Balance Sheet (Conservative Accounting)
3. Cheap Price vs Likely Value

This is different from other formulations of margin of safety, because one has to take into account factors that make it less certain that we can calculate value.  Many value managers were buying cheap financials up until September 2008, only to realize that their estimates  of value were wrong because credit losses would be far worse than expected.

Good stock analysis begins with good bond analysis.  If you wouldn’t buy a bond from the firm, you probably shouldn’t buy the stock.  Value investing is conservative, and looks for situations where there is little credit risk.

Conclusion

If you want to read  summary of my portfolio rules, you can find them here.  I am a firm believer in value investing, but I realize that there are many ways to approach the process.  I watch other value investors, and continue to learn.  Good value investors are lifelong learners, and generalists with broad knowledge.  It is not a narrow discipline, but one that can accommodate new knowledge.

Full disclosure: long SFG

One thing that came out of our “employee empowerment project” was a need to improve our equity and bond fund offerings.  At the same time, a fund manager manager [FMM] came out of the woodwork and suggested to us that we could do multiple manager funds.  They had analyzed many managers and had found some that they thought were great.

The more we thought about it, the more we thought it would be a great idea. Here’s why:

• Our own abilities to find superior managers were limited.
• A few members  of our team (including me) possessed ability to analyze what FMM would bring us.  We thought we could add value.
• We came up with a clever name, “The All Pro Funds.”
• We also thought we could add value  in changing weightings every now and then, and firing managers that we felt had become uncompetitive because they were now running too much money, had critical staff losses, or were underperforming style-specific indexes by a wide margin.
• We could increase our fee a little to pay FMM and us for the additional work entailed.

And whaddaya know?  It worked.  The portfolios in aggregate  outperformed  their indexes even after fees, and fund flows increased dramatically.  The representatives had a story to tell.  Morale improved everywhere.  We were rolling, until…

A day came where we heard from one of the underlying managers that FMM had recommended their termination because they wouldn’t rebate more of their fees back to FMM.  I would not say that we went ballistic when we heard this — instead, we went cold on FMM.  Act fast?  No, act deliberately.  The senior officers tasked me and the #2 guy in marketing to deal with the problem.

We had a rule in our division — we will pay disclosed compensation, or we will pay undisclosed compensation, but we will never pay disclosed and undisclosed compensation.  Why?  We wanted our clients to know that if compensation was disclosed, that’s all there was.  If there is no “sticker price” but you are happy with the services provided, and don’t need to know what any agent is making that’s fin with us.  But we will not pass more money quietly to those that have said, “This is the sticker price.”

FMM had violated our sense of ethics to the core.  The two of us decided to put out an RFP, asking them to bid to help manage the now $1 Billion of assets. We excluded FMM.  We chose 10 well known manager consultants. Most responded to the RFP and we invited 5 to come present to us on a given day in spring.

-=-==–=-==-=-=-=-

I need to mention one other thing.  When we first started dealing with FMM, we appreciated their qualitative research, which seemed to have some punch.  After a year, they discovered returns-based style analysis.  This allowed them to analyze many more managers just by looking at their returns, and correlating them to a variety of equity and other indexes.  They stopped the qualitative research.

The first time I saw it, I thought it was hooey, even as I think MPT is hooey.  When you have a lot of highly correlated indexes, any attempt to intuit the style of a given manager is problematic; the error bands get too wide.  It is too difficult to determine what the correct answer is.  The optimal answer mostly represents happenstance, and not fact.  Tiny tweaks to the data produce big changes in the answer.  Not a good system.

There was one incident where I met with their new quantitative analyst, a woman 10 years younger than me.  She ask if we understood how the method worked.  I replied with some mathematical jargon regarding the method, leading her to say, “Oh, so you *really* understand this.”

Also, when I analyze a manager, I like looking at what they own.  I like looking at their trades.  I want to see consistency with what they claim is their strategy.  I also like to hear why they do what they do, and what sustainable competitive advantage they think they have.  There is value in that style of analysis.  There is little value in analyzing returns.

=-=-=-===–==-==–

To our surprise, one well-known consultant [call them STAR] that had no for-profit clients was one of the five.  The leader said it was a one-time experiment, so they were evaluating us, as much as we evaluated them.

On the day when they came to present, the presentations were all over the map, from highly professional to “did not prepare.”  Some big names could not answer basic questions about what sustainable competitive advantages they brought to the process, or were fuzzy about how they earned their money.

STAR had the best presentation, services, model, ethics, etc.  It was almost “no competition,” and they liked us as well.  We hired them, much to the chagrin of FMM, who begged us to keep them.  It had the following positive results:

• Management fees down by 60%
• Fund manager fees down by 50%
• Far better marketing cachet
• Better models for investment analysis.

We reduced client fees, but had better margins, and still greater growth.  Our division was transformed thorough the two projects.  Before we started our ROE was around 8%, and we were growing AUM at a 5-10% rate.  By the time all these changes occurred, our ROE was 25%, and our growth rate was not far from that.  We were now the stars of the firm, even though the firm culturally could not acknowledge that, because the life division was so big.

I learned several things from this five-year escapade:

• Creating a desirable investment product takes work.  If you do something different that seems to add value, it will attract clients. (“We manage the managers for you, so you don’t have to”)
• Focus on ethics in those you work with.
• Reduce fees where possible, both your own, and that of suppliers.
• Name recognition helps.
• Be careful what you accept as analysis.  Just because there is clever math does not mean it represents how reality works.
• If you don’t take chances, you won’t achieve anything great.  We didn’t have to burn our old strategy, and move to multiple manager funds, but we did it, and it made clients a lot happier.  The added work was work that that we liked to do.
• Even if you have a supplier that did something good for you, do not tolerate breaches in ethics.  Find someone else to help you even if it costs more.  That it cost us less was merely a plus.

There’s kind of a rule of thumb in Asset-Liability management, that you match liquidity over the next 12 months, and match interest rate sensitivity overall.  I would do more than that, creating my own randomized interest rate models, as well as a new way of creating structured randomness in simulation models.  For a brief period of time, I had one of the best multivariate randomness programs out there, eliminating the problem of correlations in higher dimensions common with Hammersly points.  (My work was not theoretical, but intuitive… once I saw how the randomness was created, I figured out how to de-correlate the higher dimensions (since it was based on prime numbers, create more number than you need, and use a higher prime number to select observations.)

Anyway, when I brought my full-interest rate curve scenarios to the investment department in 1994, they said to me, “These are the first realistic interest rate scenarios we have ever seen.  Did you constrain them?”  I told them “No, just weak mean reversion.  Noise dominates in the short run, mean reversion dominates in the long run.”

As a result, for the lines of business over which I had oversight, we measured our interest rate mismatch in terms of days, weeks, and months, but never years.  Please ignore this incident where things drifted, but worked out exceptionally well (really, that should be a part of this series).  We published a document to show everyone how well we managed interest rate risk in Provident Mutual’s pension division.  We used scenarios far beyond what was required to show how well we did our work.  The regulators never complained.

At that point in time, the ability to integrate residential mortgage-backed securities into cash flow analyses was rudimentary at best.  But I found ways to make it work, most of the time.  That said, I remember joking with the MBS manager in late 1993, and saying there was a new term for a well-protected PAC bond.  He asked, “What is it?”  I replied, “Cash.”  He sarcastically said, “Oh, you are so funny.”   That said, I pointed out to the investment department that some of their bonds that they thought would last another four years would disappear in 2-3 months.

Then there was the floating rate guaranteed investment contract project that I eventually killed because it was impossible.  You can’t argue with expectations that are unrealistic.  Even better, I beat the Goldman Sachs representative.

In running the GIC desk at Provident Mutual, I had to review a lot of strategies because making money on short-term bonds/loans was difficult, and difficult the degree that I doubted as to whether we were in a good business.  On the bright side, I protected the firm until the day that we  could not write any more  GICs, because our credit quality was too low.  That was the fault of the less entrepreneurial part of the company, so I couldn’t so much about it, except close my operations down.  I asked the senior management team to provide a guarantee to my GICs, but they refused.

As such, I shut the line of business down.  With clients that were unreasonable over credit quality, and management unwilling to extend credit protection to GICs, the battle over GICs was ended, and I sent the line into runoff.

Five years later, as we were now part of the same firm I stood at the estate of John Dwight, with a young woman that I had sold the last GIC of Provident Mutual to, I said, “The end of the GIC business of Provident Mutual.”  We talked, she smiled, but it was part of the end of an era, because GICs were a minority of the assets in Stable Value funds.

If nothing else, this helps to highlight the impermanence of all that is done in financial firms.  I know this in my own life, but I am sure that it is true for most people in finance.

When I worked for Pacific Standard, which had the dubious distinction of being the largest life insurance insolvency of the 1980s, I had few investment-related tasks.  Investments were handled by the overly aggressive parent company Southmark, which gave little attention to risk.

But I knew things weren’t going well, and so I interviewed widely, finally landing two job offers with Midland National and AIG.   I chose the spot with AIG, because they led me to believe I would work on the international side.  When I arrived, lo, I had a job on the domestic side.  As far as the job went, had I known I would be placed on the domestic side, I would have rather gone to Midland National.  They thought I had real leadership potential — whether true or not, that’s what I was told, and I would not have minded living in South Dakota, or nearby.  As it was, there were many good things that happen to me as a result of living in-between Wilmington, Delaware, and Philadelphia, living on the PA side of the line for reasons of adoption and homeschooling.

When I got to AIG, there was one main thing that involved my risk management skills.  AIG parent wanted growth in GAAP earnings.  They wanted to see a 15% ROE, which few in the life industry were attaining.  In order to do that, they entered into reinsurance treaties (before I arrived).  These would lever up the balance sheets of the subsidiary companies, without incurring debt.  Most of them passed risk to the reinsurers, one did not.

So, when I was called into an examination by the Delaware State Insurance Department auditor over the one treaty that did not pass risk, he said to me, “You know this treaty does not pass risk.”  I replied, “Under ordinary circumstances, I would agree, but the reinsurer has taken a significant loss from this treaty.”  He said, “What do you mean?”  I replied that when Congress passed the DAC tax, the reinsurer suffered the loss — they paid up front, and we pay over time, with zero interest.

He looked at me and said that reinsurance treaties did not exist to cover tax policy, and that the treaty was a sham.  I just shrugged.  I was not the creator of the treaty, and would not have done it if I had been at AIG two years earlier.

But the there were the two larger treaties that passed risk with a vengeance to a large reinsurer [LR] who is no longer a reinsurer (if anyone wrote treaties like these, he might not be a reinsurer anymore either).  In one sense, the treaties were structured like the trading requirements in CDOs.  If you must trade:

• Get more income
• Don’t give up rating
• Don’t extend maturity
• And a few more smaller things.

I was not there when the treaties were created.  Had I been there, I would have paid a lot more attention to them, and instructed the investment department to set up segregated portfolios, which was not done.  As it was, bonds that underlay the treaty were casually sold as if free to do so.

Now I arrive on the scene.  After reading the treaties, and looking at the data, I conclude that the treaties have been abused on our side.  I suggested to LR that I go through the history, and reallocate bonds that would have fulfilled the treaties strictures, an re-work the accounting so that the terms of the treaty would be fulfilled.  Initially LR agreed to this.

The treaty passed all investment risk to the reinsurer, so defaults would hit them.  What was worse, the liabilities underlying the treaty were structured settlements.  (Structured settlements result from a court case where someone is injured.  The defendant offers to buy from a reputable life insurer an annuity that will make the requisite payments.  Low bid wins, and if the plaintiff is badly injured, the cost goes down for payments that terminate at death.  That’s where most of the bad estimates com in.)  In those days, structured settlements were a “winner’s curse.”  If you won, it was because you mis-bid.  AIG Domestic Life Companies regularly overbid for their business (as did most of the industry).  LR did not do enough due diligence to see the underwriting errors.

I did a mortality study to estimate how badly we needed to increase reserves, and lo, it was more than $100 million, all of which would flow to LR.  LR decided to sue.  After I had gone on to Provident Mutual, AIG settled with LR.  Our missteps with the assets made the case tough, and the reinsurance treaty was rescinded.  That should have been enough to jolt AIG’s earnings for a quarter, but it did not.  Funny that, and it always left me a little suspicious of AIG.  (And LR.)

Before I left AIG, I had clipped the wings of the underwriters of the structured settlements so that they could not write on cases for the most severely disabled.  I also shut down a tiny line of variable annuities that was losing money left and right to an outsourcer who had a sweet contract from a prior management team, but upon leaving AIG I did not feel that great, because I had not built anything — most of my time had been spent trying to limit losses from prior bad underwriting and planning.  It wasn’t fun, and I loved my next company more because I got to build.

PS – a prior note on AIG.

Letting the document speak, here are a few notes, starting with with the most significant part of the risk factors:

Investments are unusually concentrated and fair values are subject to loss in value.

We concentrate a high percentage of our investments in equity securities in a low number of companies and diversify our investment portfolios far less than is conventional in the insurance industry. A significant decline in the fair values of our larger investments may produce a material decline in our consolidated shareholders’ equity and our consolidated book value per share. Under certain circumstances, significant declines in the fair values of these investments may require the recognition of other than-temporary impairment losses.

A large percentage of our investments are held in our insurance companies and a decrease in the fair values of our investments could produce a large decline in statutory surplus. Our large statutory surplus serves as a competitive advantage, and a material decline could have a material adverse affect our ability to write new insurance business thus affecting our future underwriting profitability.

Buffett does very well, but I know of no other insurer that invests so much in equities funded by insurance liabilities.  There is a real risk that if the markets fall hard, a la 1929-32, 1973-4, 2007-8. that BRK would be hard-pressed, particularly if there were some significant disaster like Katrina or Sandy, or set of disasters like 2004 or 2011.

And a note on the accounting change that Buffett mentioned in his letter, but did not decide to describe:

Underwriting expenses incurred in 2012 increased $586 million (21.1%) compared with 2011. The increase was primarily the result of a change in U.S. GAAP concerning deferred policy acquisition costs (“DPAC”). DPAC represents the underwriting costs that are eligible to be capitalized and expensed as premiums are earned over the policy period. Upon adoption of the new accounting standard as of January 1, 2012, GEICO ceased deferring a large portion of its advertising costs. The new accounting standard was adopted on a prospective basis and as a result, DPAC recorded as of December 31, 2011 was amortized to expense over the remainder of the related policy periods in 2012. Policy acquisition costs related to policies written and renewed after December 31, 2011 are being deferred at lower levels than in the past. The new accounting standard for DPAC does not impact the cash basis periodic underwriting costs or our assessment of GEICO’s underwriting performance. However, the new accounting standard accelerates the timing of when certain underwriting costs are recognized in earnings. We estimate that GEICO’s underwriting expenses in 2012 would have been about $410 million less had we computed DPAC under the prior accounting standard and that, as a result, GEICO’s expense ratio (the ratio of underwriting expenses to premiums earned) in 2012 would have been less than in 2011.

The point is that BRK’s underwriting result would have been very good without the accounting change.  The accounting change was a good thing, though.  Companies trying to inflate profits look for every marketing expense that they can deem an “investment.”  All of those costs would be spread over the life of the policies, rather expensed in the current year.  The new accounting standard limits what costs can be expensed to those that are truly marginal to the business produced.

Final note: They lost money on annuity reinsurance and retro at Berkshire Hathaway Reinsurance Group [Pp 34-36].  Retro sprang from new claims.  On annuities:

The annuity business generated underwriting losses of $178 million in 2012, $118 million in 2011 and $114 million in 2010. Annuity underwriting losses reflect the periodic discount accretion of the discounted liabilities established for such contracts as well as adjustments for mortality experience.

I am not sure I would want to reinsure annuities; I’m not sure that it is possible to insure long term investment guarantees, no matter how truncated.

Full disclosure: long BRK/B

From a reader:

I only recently stumbled upon your blog, but I’m hooked.  I can’t thank you enough for taking the time to share your financial insight, experience and wisdom.

I’m a new entrant to the financial services industry (3 weeks on the job) and feeling ill-equipped without a finance degree. I’m struggling with the application of equity valuation. I’ve read several DCF valuation books and can recite all the valuation ratios, but I still have trouble looking at a companies financial statements and using them to make a judgement on a companies stock price.

Do you have any book recommendations for mastering fundamental equity valuation? 

Any help would be greatly appreciated. 

Any book by Aswath Damodaran on valuation, or Michael Mauboussin’s book on Expectations Investing will give you the theoretically correct view on how to value any sort of company.  Also, this book by James Valentine is very useful.

But I want to make your life easier.  Typically, by industry there is one simple metric that drives valuation at any point in time.  That typically gears off of the maturity of the industry, and its need for additional capital.  For those that are math nerds, there is the true model, but us lesser mortals can’t run it.  But each industry faces constraints that others don’t, and so the true model in a given industry becomes simpler to a first approximation: focus on price-to-sales, price-to-book, price-to-earnings, or the PEG ratio, among other ideas.

These simpler valuation measures focus on what is tough to do.  Can we sell more?  Can we increase our profit margin?  Can we grow our business more rapidly?

This is why sell side analysts in a given industry do not use the full valuation models listed above, but use a partial version of them, as is appropriate to their industry.

It’s useful to know the overarching model, as the above books will give you. But practically, every industry is valued differently, because each one faces different constraints, and that drives their valuations.