Recently at RM, I wrote:

David Merkel
Buy Other Insurers off of the Bad AIG News
2/12/2008 2:54 AM EST

Sometimes I think there are too many investors trading baskets of stocks, and too few doing real investing work. I have rarely been bullish on AIG… I think the last time I owned it was slightly before they added it to the DJIA, and I sold it on the day it was added. Why bearish on AIG? Isn’t it cheap? It might be; who can tell? There’s a lot buried on AIG’s balance sheet. Who can truly tell whether AIG Financial Products has its values set right? International Lease Finance? American General Finance? The long-tail casualty reserves? The value of its mortgage insurer? I’m not saying anything is wrong here, but it is a complex company, and complexity always deserves a discount.

You can read my articles from 2-3 years ago where I went through this exercise when the accounting went bad the last time, and Greenberg was shown the door. (And, judging from the scuttlebutt I hear, it has been a good thing for him. But not for AIG.)

AIG deserves to be broken up into simpler component parts that can be more easily understood and valued. Perhaps Greenberg could manage the behemoth (though I have my doubts), no one man can. There are too many disparate moving parts.

So, what would I do off of the news? Buy other insurers that have gotten hit due to senseless collateral damage (no pun intended). As I recently wrote at my blog:

If Prudential drops much further, I am buying some. With an estimated 2009 PE below 8, it would be hard to go wrong on such a high quality company. I am also hoping that Assurant drops below $53, where I will buy more. The industry fundamentals are generally favorable. Honestly, I could get juiced about Stancorp below $50, Principal, Protective, Lincoln National, Delphi Financial, Metlife… There are quality companies going on sale, and my only limit is how much I am willing to overweight the industry. Going into the energy wave in 2002, I was quadruple-weight energy. Insurance stocks are 16% of my portfolio now, which is quadruple-weight or so. This is a defensive group, with reasonable upside. I’ll keep you apprised as I make moves here.

What can I say? I like the industry’s fundamentals. These companies do not have the balance sheet issues that AIG does. I will be a buyer of some of these names on weakness.

Position: long LNC HIG AIZ

Look, back when AIG had a AAA rating, there was a reason to hold the whole thing together, because of cheap financing.  Today, AIG suffers from a conglomerate discount, because no one can understand the balance sheet.  (Can anyone inside AIG understand all the exposures that they face?)

Simpler is better.  Simple companies get better valuations, and the managers are sharper at financial controls, because they don’t have to cover as much ground.  They can focus.  So it should be for AIG, if they want to unlock value.  (Perhaps AIG is the Citigroup of the insurance industry…)

Full disclosure:  long LNC HIG AIZ

There are real advantages to having someone else do the investing for you if you are not an expert on the subject.  Pity the poor 401(k) investor who has to struggle with the right asset allocation.  He typically does not have the constitution to control his emotions in investing; fear and greed overcome many institutional investors.  What can the poor guy do?

  • He could get some advice, and then, “set it and forget it,” with occasional rebalancing.
  • He could learn about investing and try to lean against the wind.
  • He could flounder — that’s the default option.
  • Even just investing in a balanced fund is better… one fund does it all.
  • Even a money market fund might beat the guy who naively tries to time the market.

But what of a local municipality?  I’m sure some municipal Treasurers look at Florida and ask, “What if that happened here?  What would I do?  Should I try to manage the money internally?”

That’s a tough decision, and one that should be approached slowly, having discussions with all interested parties.  One main idea that I would put forth here is that it is not an on/off decision.  Let me explain.

In general, the best mergers are little ones, where a larger company acquires a smaller one that gives them:

  • Access to a new market for existing products
  • Access to a new technology that makes existing products cheaper or better
  • A new complementary product
  •  A new geographic region to sell in.
  • Etc.

The same is true of deciding to insource investment management.  It does not have to be done all at once.  Start with something simple and near to you, like cash management.  It might require the hiring of one new person with expertise, or hiring one new firm that would do it in the place of the existing relationship.  That initial move can enable later moves in other asset classes.  The idea here is to be incremental, build up expertise slowly, and analyze competitive advantage.  It may that:

  • The existing investment pool is on net better than an external manager, or what could be done internally.
  • We are large enough to hire an internal professional to manage it, earning better returns after expenses.
  • The external manager has resources that the existing pooled arrangement can’t touch, and is better than an internal arrangement can do.

The answer will differ by municipality, but the question should be asked.  Now, there are spillover benefits from having investment professionals on staff; they can be internal consultants on other investing questions.  “Should we enter into this swap agreement, together with the municipal bond we are floating?”  “What do you think the best borrowing option is for us at present?” “What do you think about the managers of our defined benefit plan?”  And there are more questions like that.  You would not think to ask them unless you had someone on staff that you thought could answer it impartially, and competently.

As for my own experience, I have been that internal consultant at many of the firms at which I have worked.  And it helped the businesses, because they trusted someone that they knew.  With that, I simply close by saying look at your existing investment relationships, and test them against a different third party relationship, and managing internally.

Getting help in investing is a tough decision.  Who is worth the money that you will pay?  Precious few.  In equities, I could probably come up with a dozen “long only” managers that have real skill, and are worth their fees with decent probability.  With hedge funds and private equity, the questions are harder, and would have a harder tme judging who has a sustainable competitive advantage.

With bond funds, the answer is simple.  Go to Vanguard.  Almost all bond managers earn roughly the same amount before fees. Over a long period of time, fees make up most of the difference in performance.  In general, low fees work with equities, but with more noise.  With index funds, the lower the fees the better, they are generic.

Now there are a few places where additional money might help.   Getting a good financial plan done can be worth the money.  For those that are wealthy, advice in limiting tax liabilities is usually worth it, though be careful when things get more complex than you can understand.  Also, insurance products can be useful, but don’t let someone sell you what is convenient for them.  Get advice from someone who won’t earn a commission, and then buy the products that you truly need.

Be careful, do your research, and buy what you want to buy.  Don’t buy what someone wants to sell you.

When I was a corporate and mortgage bond manager, I would have to look through prospectuses, if the bonds weren’t vanilla in nature. There was a division of labor — credit analysts would opine on the likelihood of whether a company was “money good,” and portfolio managers would try to decide relative value, analyze structure issues, and figure out whether the bond fit client needs.

The structure part didn’t come up often, but when it did, you’d have to read through a prospectus between a quarter of an inch to an inch thick. There were rewards to doing that. Sometimes I learned that protections weren’t what they seemed… I remember looking at an Enron privately placed bond, and after looking at the complex structure, asking what would happen if Enron’s stock price fell so hard that they couldn’t issue preferred or common equity to redeem the notes. I was told that Enron was a very successful corporation, and that wouldn’t happen. We didn’t buy the deal, and we let some of our existing Enron bonds mature. The protections might be valuable in a minor crisis, but not in a major one. In a major crisis, they would be the equivalent of unsecured debt.

After Enron blew up (and we took losses on the smaller amount of bonds still held — that’s another story), all Enron-like deals began to founder in the market. Dominion Utilities had a bond, Dominion Fiber Ventures, that was an Enron-like structure. It was only 3% of their capital structure, though, so unlike Enron, it wouldn’t kill them. We quietly bought as much as we could, after I read the prospectus, saw the protections (must issue preferred stock to redeem bonds if downgraded and stock price is below a certain price for so much time), and saw that Dominion guaranteed the debt. Dominion’s stock price did fall below the threshold, and a downgrade might come, so Dominion negotiated with bondholders to redeem the debt. We had a 10%+ gain plus interest in less than a year.

My point here is that protective measures in bonds must be adequate for the size of the issue involved, and must be capable of handling a big crisis to truly be effective. With Dominion, the protections were adequate, with Enron, they weren’t. Protective systems can work when they only have to take care of 3% of the capital structure — they will be inadequate at 50%.

Let me point you to a few other areas where this can be a problem. General American went under when they had ratings triggers on their floating rate GICs, as did ARM Financial. All it took was a downgrade, and when the money market funds exercised their puts, they couldn’t meet the redemptions, and they were insolvent. For GA, it was 25% of their capital structure. Metlife bought them for less than 75% of their net assets, and paid off the claimants. Mutual Benefit died for similar reasons. (I was a small part of trying to eliminate such triggers in property-catastrophe insurance.)

Or consider the financial guarantors. What if many different types of insured debt got into trouble at once? We may be seeing that now. If it’s not enough to see structured products in trouble, what of municipalities with soft real estate markets, like Vallejo? The rating agency models give some benefit to lack of correlation in the business mix, but in a systemic crisis, there is greater correlation. Insured obligations are AAA (or if you speak Moody’s Aaa) so long as the system is not overwhelmed. In normal times financial guarantors are money-spinners. There are few defaults, and nothing that is concentrated.

Or consider the auction rate securities markets, with all of the failed auctions of late. The dealers bought bonds when it was to their advantage, or, at least, not a big disadvantage. But when a tidal wave came, they protected themselves and not their issuers. Municipalities are working to refinance the high cost debt that they now have. The end result is bad but not horrid, but it will lead to steep yields in the long end of the muni curve for a while. (Also student loans and closed-end muni funds…)

Finally, think of the variable rate demand note (obligation/bond) market (Hi, Liz, another good article!). It is similar to the auction rate securities market, except there are banks that guarantee (for a time, sometimes to maturity, but usually for more like a year or two) to repurchase notes at par, so long as the municipality is still solvent, and the guarantor is still solvent, and not severely downgraded. The escape clauses have not been triggered, so now the banks that guaranteed repurchase at par must buy the bonds. What happens if the bank runs out of liquidity to make the purchases? The bonds will trade decidedly below par in most cases, even at the maximum interest rate payable.

I could go on from here, and talk about other protective structures that fail in disaster scenarios (Florida Hurricane Catastrophe Fund?), but you get the idea. Truth is, almost ant protective structure will fail, given a large enough crisis. Strong as the GSEs are, even they could fail in a large enough crisis, though the US government would likely stand behind senior obligations.

The important thing for fixed income investors is to evaluate the level of crisis that any protective structure/covenant might protect against, and how likely that crisis might be. During a period when many aspects of the credit markets are under threat, it’s too late to begin the analysis. Best to analyze when things are calm, and then ask the question, “What if?…”

In some ways the biggest risk that we face is unemployment risk, because the biggest asset that most people have is the stream of wages that they will earn from their jobs.

Twenty years ago, as a young actuary who had just gotten his ASA, I made a promise to myself that I would build up my investment knowledge base, and spend one hour a day improving my skills.  Why did I decide to do this?  I realized that few actuaries were good with investments (then, on this side of the Atlantic), and that most of the risks that life insurance companies faced were driven by assets, not liabilities (still true for now).  That was different than what the actuarial syllabus would lead one to believe, but nonetheless true.  I only know of one life company that failed from bad liability pricing (calculation of premiums).  All the rest died from bad asset strategies.

That “one hour a day” (six days a week) made me invaluable to many of the companies that I served, and opened a lot of employment doors for me.  It also allowed me to make a jump out of the insurance world, at least for now.  In a knowledge based economy, continually improving your skills is a great way to advance your career, and limit downside when the inevitable bumps happen due to M&A, etc.

Now, to the average person entering the work force, it pays to look at the underlying economics of the industry to see how stable employment prospects will be.  No one is perfect in making these judgments, but there are often industries to avoid.  Examples: certain traditional media companies are being destroyed by the internet.  During the tech bubble, it was cool to work for tech companies, but how much future is there if they don’t make any money.  Wall Street is wonderful, but periodic layoffs can knock out a lot of people on the margins of the business.

Also, understanding the underlying economics of your industry instantly makes you more valuable to your employer, since many only understand the technical craft that they pursue.

Finally, cultivate friends.  Be competent, but be warm.  Help others in need when they are looking for work.  Be willing to lend a sympathetic ear to colleagues in their job troubles.  Network at industry functions.  Start a blog to demonstrate expertise.  (Okay, nix that one. 😉 )  Treat vendors with kindness and respect; learn their business if you can.  Join industry task forces to solve larger problems.

We can’t control our employment futures in entire, but we can influence how well we bounce when things don’t go right. To repeat, three ways to mitigate unemployment risk:

  • Continually improve your skills
  • Understand your industry
  • Build a network of friends in your industry

One of the troubles with the way that academic research in the social (and biological) sciences is set up, is that there is a bias toward publishing research that is statistically significant. Here are some of the problems:

  1. If honestly done, there is value in publishing research that says there doesn’t seem to be any relationship between variable being studied and the cofactors. If nothing else, it would tell future researchers that that avenue has been checked already. Try another idea.
  2. It encourages quiet specification searches, where the researcher tries out a number of different variables or functional forms, until he gets one with significant t-coefficients. Try enough models, one will eventually hit the 95% significance threshold.
  3. What is statistically significant is sometimes not really significant. The result might be statistically significantly different than the null hypothesis, but be so small that it lacks real significance. I.e., learning that a compound increases cancer risk by one billionth should not be significant enough to merit attention.
  4. Researchers are people just like you and me, and all of the foibles of behavioral finance apply to them. They want tenure, promotions, don’t want to be let go, respect from colleagues and students, etc. They have biases in the selection of research and the framing of hypotheses. For example, we can’t assume that stock price movements have infinite variance, because then Black-Scholes, and many other option formulas don’t work. The Normal distribution and its close cousins become a crutch that allows for papers to get published.
  5. Once an idea becomes a researcher’s “baby”, they tend to nurture it until a lot of contrary evidence comes in. (I’ve seen it.)
  6. Famous researchers tend to get more slack than those that are not well-known. I would trot out as my example here returns-based style analysis, which was proposed by William Sharpe. When I ran into it, one of the first things I noticed was that there were no error bounds on the calculations, and that the cofactors were all highly correlated with each other. The paper didn’t get much traction in the academic world, but was an instant hit in the manager selection consultant community. A FAJ paper in 1998 (I think) came up with approximate error bounds, and proved it useless, but it is still used by some consultants today. (I have many stories on that one; it is that only time that I wrote a pseudo-academic paper in my career to keep some overly slick consultants from bamboozling my bosses.)
  7. Data sets are usually smaller than one would like, and the collection of raw data is expensive. Sample sizes can get so small that relying on the results of subsamples for various cofactors can be unreliable. This is a particular problem in the media when they publish the summary results on drug trials, but don’t catch how small the samples were. People get excited over results that may very well get overturned in the next study.
  8. Often companies fund research, and they have an interest in the results. That can bias things two ways: a) A drug company wants their proposed drug approved by the FDA. A researcher finding borderline results could be incented to look a little harder in order to get the result his patron is looking for. b) A finance professor could stumble across a new profitable anomaly to trade on. That paper ends up not getting published, and he goes to work for a major hedge fund.
  9. The same can be true of government-funded research. Subtle pressure can be brought on researchers to adjust their views. Politically motivated economists can ignore a lot of relevant data while serving their masters, and this is true on the right and the left.

The reason that I write this is not to denigrate academic research; I use it in my investing, but I try to be careful about what I accept.

Now, recently, I took a little heat for making a comment that I thought that the unadjusted CPI or median CPI was a better predictor of the unadjusted CPI than the “core” CPI. So, I went over to the database at FRED (St. Louis Fed), and downloaded the three series. I regressed six month lagged unadjusted, median, and core CPI data on unadjusted CPI data for the next six months. I made sure that the data periods were non-overlapping, and long enough that data corrections would induce little bias. I constrained the weights on my three independent variables to sum to one, since that I am trying to figure out which one gets the most weight. My data set had 80 non-overlapping six-month observations stretching back to 1967. Well, here are the results:

  • Intercept: -0.0002 (good, it should be close to zero)
  • Unadjusted CPI: 0.1720 (prob-value 12.3%)
  • “Core” CPI: -0.1665 (prob-value 11.2%)
  • Median CPI: 0.9945 (no prob-value because of the constraint imposed)
  • Prob-value on the F-test: 24.3% (ouch)
  • Adjusted R-squared: 1.10%. (double ouch)

What does this tell me? Not much. The regression as a whole is not significant at a 95% level. Does the median CPI (from the Cleveland Fed) better predict the unadjusted CPI than the “core” or unadjusted CPI? Maybe, but with these results, who can tell? It is fair to say that core CPI does not possess any special ability to forecast unadjusted CPI over a six-month horizon.

From basic statistics, we already know that the median is a more robust estimator of central tendency than the mean, when the underlying distribution is not known. We also know that tossing out data (“core”) arbitrarily because they are more volatile (and higher) than the other components will not necessarily estimate central tendency better. Instead, it may bias the estimate.

So, be wary of the received opinion of economists that are in the public view. Our ability to use past inflation measures to predict future inflation measures is poor at best, and “core” measures don’t help in the explanation.

1) The blog was out of commission most of Saturday and Sunday, for anyone who was wondering what happened. From my hosting provider:

We experienced a service interruption affecting the Netfirms corporate websites and some of our customer hosted websites and e-mail services.

During scheduled power maintenance at our Data Centre on Saturday Feb. 23 at approximately 10:30 AM ET, the building’s backup generator system unexpectedly failed, impacting network connectivity. This affected several Internet and Hosting Providers, including Netfirms.

Ouch. Reliability is down to two nines at best for 2008. What a freak mishap.

2) Thanks to Bill Rempel for his comments on my PEG ratio piece. I did not have access to backtesting software, but now I do. I didn’t realize how much was available for free out on the web. He comes up with an interesting result, worthy of further investigation. My main result was that PEG ratio hurdles are consistent with a DDM framework within certain moderate values of P/E and discount rates. Thanks also to Josh Stern for his comments.

3) I posted a set of questions on Technical Analysis over at RealMoney, and invited the technicians to comment.

David Merkel
Professionals are Overrated on Fundamental Analysis
2/21/2008 5:19 PM EST

I’m not here to spit at technicians. I have used my own version of technical analysis in bond trading; it can work if done right. But the same thing is true of fundamental investors, including professionals. There are very few professional investors that are capable of delivering above average returns over a long period of time. Part of it is that there are a lot of clever people in the game, and that raises the bar.

But I have known many good amateur investors that do nothing but fundamental analysis, and beat the pros. Why? 1) They can take positions in companies that are too small for the big guys to consider. 2) They can buy and hold. There is no pressure to kick out a position that is temporarily underperforming. With so many quantitative investors managing money to short time horizons, it is a real advantage to be able to invest to longer horizons amid the short-term volatility. 3) They can buy shares in companies that have been trashed, without the “looks that colleagues give you” when you propose a name that is down over 50% in the past year, even though the fundamentals haven’t deteriorated that much. 4) Individual investors avoid the “groupthink” of many professionals. 5) Individual investors can incorporate momentum into their investing without “getting funny looks from colleagues.” (A bow in the direction of technical analysis.)

When I first came to RM 4.4 years ago, I asked a question of the technicians, and, I received no response. I do have two questions for the technicians on the site, not meant to provoke a fundy/technician argument, but just to get opinions on how they view technical analysis. If one of the technicians wants to take me up on this, I’ll post the questions — hey, maybe RM would want to do a 360 on them if we get enough participation. Let me know.

Position: none

David Merkel
The Two Questions on Technical Analysis
2/22/2008 12:15 AM EST

I received some e-mails from readers asking me to post the questions that I mentioned in the CC after the close of business yesterday. Again, I’m not trying to start an argument between fundies and techies. I just want to hear the opinions of the technicians. Anyway, here goes: 1) Is there one overarching theory of technical analysis that all of the popular methods are applications of, or are there many differing forms of technical analysis that compete against each other for validity (and hopefully, profits)? If there is one overarching method, who has expressed it best? (What book do I buy to learn the theory?)

2) In quantitative investing circles, it is well known (and Eddy has written about it recently for us) that momentum works in the short run, and is often one of the most powerful return anomalies in the market. Is being a good technician just another way of trying to decide when to jump onto assets with positive price momentum for short periods of time? Can I equate technical analysis with buying momentum?

To any of you that answer, I thank you. If we get enough answers, maybe the editors will want to do a 360.

Position: none

I kinda thought this might happen, but I received zero public responses. I did receive one thoughtful private response, but I was asked to keep it private. Suffice it to say that some in TA think there is a difference between TA and chart-reading.

As for me, though I have sometimes been critical of TA, and sometimes less than cautious in my words, my guesses at the two questions are: 1) There is no common underlying theory to all TA, there are a variety of competing theories. 2) Most chart-readers are momentum players, as are most growth investors. Some TA practitioners do try to profit from turning points, but they seem to be a minority.

I’m not saying TA doesn’t work, because I have my own variations on it that I have applied mainly to bond investing. But I’m not sure how one would test if TA in general does or doesn’t work, because there may not be a commonly accepted definition of what TA would say on any specific situation.

4) One more note from RM today:

David Merkel
Just in Case
2/25/2008 4:20 PM EST

Um, after reading this article at the Financial Times, I thought it would be a good idea for me to point readers to my article that explained the 2005 Correlation Crisis. Odds are getting higher that we get a repeat. What would trigger the crisis? A rapid decline in creditworthiness for a minority of companies whose debts are referenced in the relevant credit indexes, while the rest of the companies have little decline in creditworthiness. One or two surprise defaults would really be gruesome.

Just something to watch out for, as if we don’t have enough going wrong in our debt markets now. I bumped into some my old RM articles and CC comments from 2005, and the problems that I described then are happening now.

Position: none, and there are times when I would prefer not being right. This is one of them. Few win in a bust.

There are situations that are micro-stable and macro-unstable, and await some force to come along and give it a push, knocking it out of its zone of micro-stability, and into a new regime of instability. When you write about situations like that before the fact, it is quite possible that you can end up wrong for a long time. I wrote for several years as RM about overleveraging credit, mis-hedging, yield-seeking, over-investment in residential real estate (May 2005), subprime lending (November 2006), quantitative strategies gone awry, etc. The important thing is not to put a time on the prediction because it gives a false message to readers. One can see the bubble forming, but figuring out when cash flow will be insufficient to keep the bubble financed is desperately hard.

5) This brings up another point. It’s not enough to know that an investment will eventually yield a certain outcome, for example, that a distressed tranche of an ABS deal will eventually pay off at par. One also has to understand whether an investor can handle the financing risks before receiving the eventual payoff. Will your prime broker continue to finance you on favorable terms? Will your regulator force you to put up more capital against the position? Will your investors hang around for the eventual payoff, or will they desert you, and turn you into a forced seller? Can your performance survive an asset that might be a dud for some time?

This is why the price path to the eventual payoff matters. It shakes out the weak holders, and moves assets that should be financed by equity onto strong balance sheets. It’s also a reason to be careful with your own balance sheet during boom times, and in the beginning and middle of financial crises — don’t overextend your positions, because you can’t tell how long or deep the crisis might be.

6) I agree with Caroline Baum; I don’t think that the FOMC is pushing on a string. The monetary aggregates are moving up, and nominal GDP will as well… it just takes time. The yield curve has enough slope to benefit banks that don’t face a lot of credit problems… and the yield curve will steepen further from here, particularly if the expected nadir of Fed funds drops below 2%. Now, will real GDP begin to pick up steam? Not sure, the real question is how much inflation the Fed is willing to accept in the short run as they try to reflate.

7) Now, inflation seems to be rising globally. At this point in the cycle, the FOMC is ahead of almost all major central banks in loosening policy. I think that is baked into the US dollar at present, so unless the FOMC gets even more ahead, the US Dollar should tread water here. Eventually inflation elsewhere will get imported into the US. It’s just a matter of time. That’s why I like TIPS here; eventually the level of inflation passing through the CPI will be reflected in implied inflation rates.

8 ) Okay, MBIA will split in 5 years? That is probably enough time to strike deals with most everyone that they wrote coverage for structured products, assuming the losses are not so severe that the entire holding company is imperiled. If it’s five years away, splitting is a possibility, but then are the rating agencies willing to wait that long? S&P showed that they are willing to wait today. Moody’s will probably go along, but for how long?

9) I found it interesting that AQR Capital has not been doing well in 2008. When quant funds did badly in the latter half of 2007, I suffered along with them. At present, I am certainly not suffering, but it seems that the quants are. I wonder what is different now? I suspect that there is too much money chasing the anomalies that the quant funds target, and we reached the end of the positive self-reinforcing cycle around mid-year 2007; since then, we have been in a negative self-reinforcing cycle, with clients pulling money, and the ability to carry positions shrinking.

10) Now some graphs tell a story. Sometimes the story is distorted. This graph of the spread on Fannie Mae MBS is an example. Not all of the spread is due to the creditworthiness of Fannie Mae. Those spreads have widened 30 basis points or so over the past six months for Fannie’s on-the-run 5-year corporate bond, versus 50 basis points on the graph that I referenced. So what’s the difference? Increased market volatility makes residential MBS buyers more skittish, and they demand a higher yield for bearing the negative optionality inherent in RMBS. Fannie and Freddie are facing harder times from the guarantees that they have written, and the credit difficulties at the mortgage insurers, but it would be difficult to imagine the US Government allowing Fannie or Freddie to default on senior obligations.

That’s another reason why I like agency-backed RMBS here. You’re getting paid a decent spread to bear the risks involved.

11) I would be cautious about using prics from CMBX, ABX, etc., to make judgments about the cash bonds that they reference. It is relatively difficult to borrow and short small ABS and CMBS tranches. It is comparatively easy to buy protection on the indexes, the only question is what level does it take to induce another market participant to sell protection to you. When there is a lot of pressure to short, prices overshoot on the downside, and stay well below where the cash bonds would trade.

12) One last point, this one coming via one of our dedicated readers passing on this blurb from David Rosenberg at Merrill Lynch:

A client sent this to us last week

It was a New York Times article by Louis Uchitelle in December 1990 on the housing and credit crunch. In the article, there is a quote that goes like this – “This is different from the experience of the Great Depression, but something related to the 1930’s is beginning to happen”. Guess who it was that said that (answer is at the bottom of the Tidbits).

Answer to question above

Ben Bernanke, a Princeton University Economist” (and future Fed chairman, but who knew that then?).

My take: it is a very unusual time to have a man as Fed Chairman who is a wonk about the Great Depression. That makes him far more likely to ease. The real question is what the FOMC will do if economic weakness persists, and inflation continues to creep up. I know that they want to save the day, and then remove all policy accomodation, but that’s a pretty difficult trick to achieve. In this scenario, I don’t think the gambit will work; we will likely end up with a higher rate of price inflation.

This piece is a work in progress, so I solicit your feedback on it. How could it be improved?


I enjoy it when my expectations are proven wrong, because it means that I learned something in the process. When I began preparation for this post (which will probably have two parts, because I am having difficulty posting files, tables and pictures at my blog), I expected to write a post that would conclude that the PEG ratio (P/E divided by the anticipated growth rate expressed as an integer) is a nifty market artifact, but had no sound theoretical grounding.

The answer to the question in my title is complex. The answers are No, Sometimes, and Yes.

  • If you’re a deep value investor: No.
  • If you’re a moderate value or core investor: Sometimes.
  • If you’re a fundamentally-driven moderate growth investor: Yes.
  • If you’re an aggressive growth investor: No.

When I did my earlier post on my version of the Fed Model, I began by showing that it was a simplification of the simple version of the dividend discount model [DDM], which states that the value of a stock is equal to the present value of its future dividends. I’m going to do the same thing here with a few changes:

· I can’t prove what I am stating analytically, that is, by manipulating equations. I’m going to do it through scenario analysis and regression.

· My piece on my Fed model used the simple DDM. This piece uses a three-stage DDM. The stages are growth, transition, and maturity. For those with access to a Bloomberg Terminal, my implementation is a more conservative version of what they did.

Three-Stage DDM Assumptions

  • · Initial forecast earnings (E1)
  • · Initial dividend payout ratio as a portion of earnings (PR1)
  • · Growth rate of earnings in the first phase of the model (g)
  • · Length of the first phase (5 years)
  • · Length of the second transition phase (6 years)
  • · Ultimate earnings growth rate in maturity (6%)
  • · Ultimate payout ratio in maturity (50%)
  • · Discount rate for the dividend stream (ks), otherwise known as the required rate of return (i.e., what does an investor have to expect to earn in order to get him to part with his cash?)

In brief, in the first phase of the model, earnings grow at a rapid rate, and dividends are paid at a relatively low rate, in the second (transition) phase, the earnings growth and dividend payout rates grade linearly into the rates of the ultimate phase. The resulting dividend stream gets discounted at a discount rate reflecting the riskiness of the company.

Limitations of the Model

  • · It is difficult to forecast earnings for next year, much less give a growth rate for the next 5 years. I use sell side estimates as an initial jumping off point.
  • · Companies grow erratically.
  • · The maturation of a company is rarely so linear.
  • · The lengths of the first two phases are somewhat arbitrary, though the sell side typically does 5-year growth rates.
  • · A 6% growth rate in maturity is consistent with long term nominal GDP growth, but it is still quite an assumption.
  • · Payout rates and growth rates should be inversely correlated. To the extent that capital constrains business growth, a higher rate of dividend payout should result in a lower earnings growth rate.
  • · The discount rate is difficult to calculate. Theoretically, it should be 2-3% percent higher than the highest yield on the longest, most subordinated debt or preferred of the company. If a company has no debt, compare it to the yields of bonds of other companies with similar put option implied volatility 20% or more out of the money. Then add 2-3% to those yields.
  • · Payout rates and the discount rate should be negatively correlated. Companies with high payout rates will be judged to be less risky most of the time, and vice-versa.

All that said, the DDM is a model, and a richer model than the PEG ratio. My question became, “Are there conditions where the results of the DDM resemble a PEG ratio?” The answer to that is yes, when:

  • The discount rate is 14% or lower
  • At lower discount rates, only for higher P/Es. For example at a discount rate of 8%, the PEG ratio works for P/Es 16 and higher.

Now I oversimplified my conclusions here. Look at this graphic:

Validity Space
Or, based off of that, consider this graph, which shows the PEG hurdle rates as a function of initial P/Es. and cost of capital (discount) rates:

Price-to-Value Graph

How did I come to this result? For differing levels of the discount rate, I varied P/E levels, calculating the initial phase growth rate that would make price equal to value in the DDM. Those P/E and growth levels gave me the PEG ratios. Those PEG ratios were often quite flat for higher P/Es at a given level of the discount rate of 14% or below. There is usually a bit of a smile or smirk, but you can see an average level.

At 16% or higher levels of the discount rate, the PEG ratio falls apart. At low levels of P/E the required PEG ratio should be low. At high levels of P/E, the required PEG ratio can be higher. The intuition here is that situations with high discount rates, and thus high risk, require high growth to fuel value in a DDM calculation.

At low discount rates, and low P/Es, the DDM says that value investors don’t need much growth at all in order to buy good values. If one considers the inverse of the P/E, the E/P, or earnings yield, when it is greater than the discount rate, it is hard to lose money, even when earnings don’t grow. Even more so when the dividend yield exceeds or is near the discount rate.

A Formula for the PEG Ratio Hurdle

Taking the average PEG hurdle rates for P/Es 16 and above, where price equaled DDM Value, for various discount and payout rates, I calculated a regression to give a more general PEG hurdle rate formula. The factors appeared multiplicative, so I used a formula that looked like this:

ln ( average PEG hurdle) = a + b * ln(discount rate) + c * ln(payout rate) + e (error term)

The regression had an adjusted R-squared of 98%, with all coefficients statistically significant at prob-values of 99% or better. a was 7.8646, b was -1.3169 and c was .0752. In summary form, the formula looks like this:

Average PEG hurdle = 26.03 * discount rate-1.3169 * payout rate0.0752

Pretty good, but after a little while, I asked if I could create a formula that better represented the curves in graph 2. So, I ran the following regression:

ln (PEG hurdle) = a + b * ln(discount rate) + c * ln(payout rate) + d * ln(P/E) + e (error term)

I had a debate as to how to censor the data. I threw out data points with negative PEG hurdles in the first analysis. In the second one, I threw out negative PEG hurdles, and PEG hurdles over 2.0x. On the second analysis, my reasoning was that if PEG hurdles over 2.0 are acceptable, we’re in weird times. Now perhaps that pre-judges the situation, but the right functional form for graph 2 eludes me here. Personally, I would use the second formula here:

Formula 1: Average PEG hurdle = 0.01823 * discount rate-1.6279 * payout rate0.1039 * PE Ratio0.1893

Formula 2: Average PEG hurdle = 0.02035 * discount rate-1.4215 * payout rate0.0941 * PE Ratio0.2704

Formula 1 has an R-squared of 76%, and with 2 it is 88%. The t-statistics are all significant at 99% levels.

Now, suppose I am a growth investor and I decide to apply formula 2. I look for stocks with PE ratios of around 20, my discount rate is 15%, and the dividend payout rate is around 10%. What annual earnings growth should I be looking for over the next 5 years? The formula says 36.6%. Pretty aggressive. At a discount rate of 12%, the growth rate drops to 26.6%.

What this points out in a way is the difficulty of making consistent money in growth stocks. The earnings growth rates needed to make money in excess of the discount rate on average over time is higher than most growth investors realize.

Growth investors overpay for growth. That is one of the reasons that I am a value investor.

One final note: Jim Cramer has a limit for what he is willing to pay for growth stocks – a PEG ratio of 2.0x. Now, he’s a bright guy, so there are two ways that I can interpret this. 1) Since momentum plays a large role in Cramer’s investing, the 2.0x ceiling limits his risk while he plays momentum. Or, 2) he has longer periods of competitive advantage and transition than I do. I favor the first interpretation, because it is rare in my opinion that growth investors should pay over 1.5 times the growth rate for any investment, unless the barriers to entry are significant.


PEG ratios work for core and growth investors, but the PEG ratio hurdles needed for investment are lower than most investors think, so long as the expected rate of return (discount rate) is high.  As for me, I will stick with value investing, where the need for earnings growth is negligible.

I don’t really have a dog in the fight regarding the financial guarantors, but after reading Bill Ackman’s proposal for how to split them, I had two reactions:

  1. That’s a good idea, and
  2. You are talking your own book.

As I have said before, most companies in financial trouble would just love to split themselves in two.  Create company A with the good business, and company B with the bad business, and the holding company owns them both.  Send company B into insolvency, and the holding company hasn’t lost much… most of the value was in company A.  Who lost, though?  The creditors of the company before the split, who now rely on company B.  In the real world, it gets called fraudulent conveyance.

Ackman’s proposal avoids that.  It makes the CDO guarantor the owner of the municipal guarantor, and the holding company only directly owns the CDO guarantor.  From my example above, it would mean that the holding company would own company B, which in turn owns company A.  If B goes bust, creditors of B still have the advantage of being to draw on the value of company A in insolvency.  In this split-up, in the short run, no one’s rights are compromised.

But for Ackman, it is still talking his book, because he is trying to protect against a split-up where the holding company owns both A and B independently, because the holding company (which he is short) is worth more if the regulators allow such a split-up, and the court cases fail that challenge such a split-up.

This is one of those cases where the proposer of the idea gets ignored because of his self-interest.

It seems that a number of Constant Proportion Debt Obligations are being downgraded or forced to delever.  This was something I thought would happen; it was only a question of when.  It’s a pity that S&P did not totally abandon its model framework for CPDOs; it is less liberal now, but not consistent with the way they rate other investments.  Here’s a trip through my thoughts on CPDOs over the last 16 months:

David Merkel
Having A Sense Of Wonder
11/7/2006 2:09 AM EST

Periodically, I gain a sense of wonder from the derivative markets. This stems from the optimism of the markets vs. my knowledge of economic history. There are risks being taken that have not worked out in the past. My current wonder-generator is the CPDO [Constant Proportion Debt Obligation] market. With a CPDO, you leverage up a basket of investment grade credits, in an effort to earn a certain amount over the life of the CPDO. {Note: the CPDO is rated AAA, but the average of the underlying credits is rated weak single-A at best.

If the deal goes well, i.e. no defaults, it delivers early, and risk decreases. If defaults occur, the structure levers up more in an effort to make back what has been lost, up to a 15x leverage limit. After that, the CPDO rapidly takes on losses.

This structure is notable, because it attempts to achieve risk reduction for free, the same way the stock managers tried to do so in the mid-80s with dynamic portfolio management. It has no external guarantors, nor subordination.

The rumor at present is that these new CPDOs are leading to a tightening in the credit default swap market. Spreads are tight as a drum, so I can see the effect, if true.

Position: None, but I always get concerned when market players try to get risk control for free. Off-loading risk is never free on average.

David Merkel
Call It Complacency
11/7/2006 3:58 PM EST

Be sure and look at Tony’s blog post, “Default Insurance Costs at New Low.” I checked the other Dow Jones CDX North America Investment Grade Indexes, and yes, they also are at all time tight levels. Tony cites the spread from the newest one. Should we be worried? A little. As I noted in my post from this morning, some of this tightness is due to the CPDO market. They have to suck in a lot of long credit exposure to issue these, which puts downward pressure on spreads.

But bottoms in the stock market are an event. Tops are a process. Credit spreads are tight for long periods during the bull phase, and very fat for short periods during the bear phase. (Can I have BBB spreads in the 400s again, please?) Same for implied volatility… the VIX spikes during equity and credit market panics, but lolls around at low levels during the bull phase. This is complacency.

Trouble is, complacency can last a looong time, and many fixed income and equity managers don’t have the luxury of saying, “I think I’ll just stay in T-bills for now.” The greed of those they invest for (or their actuarial funding targets) force them into risk, often at bad times. The good times end when cash flow is insufficient to refinance marginal assets. Typically that’s three years after the issuance of debt deals that should never have been done, but in this environment, there is so much private equity amd vulture capital around that I don’t see many troubled assets not getting financing.

The party will continue a while longer. Oh look, there are the hedge fund-of-funds at the head of the Conga line, followed by the CDO equity managers, the investment banks, the credit hedge funds, and the cash bond market at the tail. What a party!

PS — I think it is irresponsible of the rating agencies to assign AAA ratings to securities like these CPDOs that are composed of BBB and single-A paper, and do not have any guarantor or subordination to protect the creditworthiness. This is akin to thinking that a martingale method, like doubling down, will protect you from loss in Vegas. It might most of the time, but you lose big when it doesn’t work.

Position: none

David Merkel
More Information on CPDOs
11/9/2006 12:25 AM EST

I’ve gotten numerous pings since my initial posts on CPDOs [Constant Proportion Debt Obligations]. This post is designed to correct a few errors, and explain how we as equity investors might profit from a potential disaster here. My first posts were based off what I read in a few blogs. They got a few things wrong, so I am correcting what I wrote. The structure levers up investment grade credit fifteen times, allowing the purchaser to buy a bond with a coupon two percent (or so) higher than Treasuries, with a AAA rating. What a deal; it is difficult to find AAA bonds yielding 0.5% more than Treasuries. (Ignoring odd beasts like CMBS IOs, etc.) I have seen reports that $1.0-1.5 billion of these have been created in the recent past, which means around $20 billion of credit exposure has been absorbed, depressing credit spreads over the last month.

I suggested in my earlier writings that the structures could only allow for 15x leverage, but they can go higher if the deals go badly at first. They only unwind and take losses if the market value of the underlying assets would drop down to a threshold level, say, around 90%-94% of par. That’s not to say that losses are limited to 6%-10%. The losses could be worse if the market is moving against them as they liquidate.

Now, how to profit? There will be some sort of crisis from CPDOs; after all, the buying in order to establish these securities has been characterized by some as a panic. At some point, there will be a situation where there is a default on one or more of the companies in one of the CPDOs. If it is severe enough, at that time, the CPDOs will have to deliver, and that will push credit spreads wider, and stock prices lower.

Since the companies involved are all big capitalization companies, we can watch the price and volume patterns on the S&P 500 Spyder, and look for where volume is cresting, while price is trough-ing, and take a long position after the crisis. Watch the VIX. When it spikes in a situation like this, there will be profits from going long equity exposure.

If it means anything, I used strategies like this in 2001-2002 to generate profits when things were going crazy. These strategies will work again when the CPDOs fail. I can’t say they will fail soon; I just know they will fail, as Dynamic Portfolio Management did in 1987.

Position: None

David Merkel
Dominion & CPDOs
12/20/2006 12:47 AM EST

I’m not alone on not liking what Moody’s and S&P have done on constant proportion debt obligations [CPDOs]. Now a rival rating agency, Dominion, better known for rating Canadian debt, has weighed in on the issue with skepticism. I’m annoyed at the irresponsibility of Moody’s and S&P for two main reasons. A weak single-A, strong-BBB portfolio should have credit losses of 10-15 basis points per year on average. Unfortunately, losses tend to come in heaps for investment grade corporate debt. No losses for five years, and large losses for two years. Now these structures are levered up 9-15 times on average, so during the two loss years, we are talking about 9-10% losses of equity over a two year period. If that is not bad enough, spreads will widen during the loss period even on healthy debt, further adding to the problems.

In the old days, say, two years ago, Moody’s and S&P would have called a CPDO structure AAA once it had de-levered, not on the prospect that it is very likely to de-lever. Remember a AAA means it can survive the Great Depression, and pay principal and interest on a timely basis. I can say with certainty that a levered portfolio of weak single-A bonds can’t do what an unlevered AAA bond can do in a period of severe economic stress.

Can rating agencies be sued for malpractice? Perhaps the boards of Moody’s and S&P should spruce up their D&O coverage…

Position: none

Beyond these pieces, I had three posts here that followed the decline:

Speculation Away From Subprime, Compendium

Stressing Credit Stress

Ten Notes on Our Crazy Credit Markets

Now we may have an opportunity as some CPDOs are forced to delever, credit spreads are being forced higher.  I commented before (all too recently) that it was time to dip our toes into the waters of credit, and buy 25% of a full position, with carefully selected credits.  I think it is now time to raise that allocation to 50%.  It is time to begin taking some credit risks; spreads are discounting a lot of unfavorable future news, and it is time to take advantage of it.  Is the current news gloomy?  You bet, and I can tell you that at the end of many days in mid-2002, I would hold my head in my hands in disbelief at the carnage.  But good credit investors must invest when the spreads are wide, and give up income when spreads are tight.

As for the 2002 carnage, I sent Cramer e-mails on the bond market back then, and this CC post recounts one of them, where Cramer used one of my e-mails for a post (he did that twice in 2002):

David Merkel
Cycling Through Cycles
2/1/05 2:54 PM ET
If you haven’t read it yet, please read Cody’s piece, “The Nature of Feedback Loops.” I do a lot with this for two reasons. First my investment methods lead me to rotate sectors, and all mature businesses are cyclical; they just aren’t all on the same cycle. Second, the insurance industry is very cyclical. I spend most of my time analyzing trends in pricing power.At cycle peaks and troughs, I tend to stop looking at quantitative data, and look for anomalous behavior that might hint that the cycle is changing. No one rings a bell at the top or bottom, and you can never get tops and bottoms exactly, but sometimes people behave funny near turning points. Greed and fear get excessive, and then people do foolish things. As an aside, before I wrote for RealMoney, I would drop Jim Cramer notes on the corporate bond market. This article resulted from one of my missives. There was still five months of craziness remaining, but I kept my trading discipline in 2002, though I sometimes wondered if I was sane. At the turns in July and October, some of my best brokers called me in a panic, saying that there were no bids in the market and many sellers. Implied equity volatility had gone through the roof.

What to do? I got out the liquidity that I reserved for such occasions and put out lowball bids for medium-quality bonds. By the time I used up all my liquidity in the early afternoon, the market had turned. Nerve-wracking, but it really made for good performance in a horrible year.

For more of my thoughts on applying cycles to investing, you can read my piece, “Evolution of an Investment Style.”


Anyway, buying credit now is a “pain trade.”  It is time to selectively take advantage of wide spreads if your investment mandate allows for it.