Philip Fisher's Investment Method

Philip Fisher, author of the classic book such Common Stocks and Uncommon Profits was known for his successful career investing in growth stocks and for his "scuttlebutt" approach to investing, which is not unlike what Peter Lynch used to do when he managed Fidelity's Magellan fund.

In chapter 10 of Common Stocks and Uncommon Profits Fisher explains how he went about making investment decisions. Here's a summary. First, he needed an idea. About one-fifth of his ideas came from investment houses, analysts, businessmen and scientists who happen to know about a particular industry. The other four-fifths came from reputable and respected professional investors themselves. Then, with the idea in mind, he did not pour over past annual reports and spends hours on detailed financial data. He simply checked the balance sheet and took a first look at the financials, to make sure the company was sane. He also did not talk to management at this point. He did instead go out doing his "scuttlebutt" method.

The "scuttlebutt" method consists of calling his network of business executive-scientists, calling competitors, suppliers, former employees, and customers of the company in question. He asked questions and used the answers to form a picture about the company and its future, often cross-referencing pieces of information from different sources, to validate its veracity. He found out how well-managed the company was, what its competitive advantages were and how well it was respected by competitors, customers and suppliers. If Fisher could not get at least 50% of his required investment criteria at this point, he would give up and move on to something else.

Only when he had at least 50% (and typically, much more) of the information needed to form a positive view on a company was when he would call the company, talk to management and visit factories and assembly lines. Typically, at that stage, about one in two companies he visited he actually invested in. But that only happened if he had enough positive information from his "scuttlebutt" approach first.

When asked whether retail investors should be expected to devote that amount of time and energy to investing, Fisher would say "absolutely" and ask the following question: "In what other line of activity could you put $10,000 in one year and ten years later be able to have an asset worth from $40,000 to $150,000? Is it either logical or reasonable that anyone could do this with an effort no harder than reading a few simply worded brokers' free circulars in the comfort of an armchair one evening a week?"

In a future post, I will talk about Fisher's 15-point investment criteria and discuss some of his other insights.


Is The Recession Over?

I don't know and I think those who attempt to answer this question in real-time are bound to look foolish sooner or later. But as a group, that's what security analysts might be getting at: a consensus opinion that the stock market recession is over.

Here's the empirical data: today, thirty two companies for which there are analyst estimates posted their quarterly earnings. Out of those, a solid 28 beat expectations, 2 were inline and 2 missed. That's an 87.5% "success" rate.

By my account, the stock market has been rallying since March due to the same reason. So, from a pure interpretation of analysts expectations, the stock market, which is a forward-looking market, might have discounted expectations too much again. And hence why we observed yet another rally today and most of the past week. This will happen until expectations become high again and companies start to occasionally miss them, the market collapses and the cycle repeats.

When we look back at today and look at a timeline of the past N recessions (on those gray-shaded graphs indicating recessions where the stock market usually leads by six to nine months), we'll know for sure if the recession was really over six to nine months from now.

What does this mean for investors? Sadly, that those who didn't invest in the recent panics of November and March might have missed some pretty good opportunities. The good news? I can't predict the market, but here's one prediction I can make: there will be other buying opportunities in the future, however distant it may be.

My advice: pick the winning companies now, assess their intrinsic value and be ready to buy at the next opportunity. We don't know how soon it could be.


The Myth of Dividend Payout Ratio Part II: Is Lower Necessarily Better?

In a previous post I commented how an insistence on low payout ratios can be harmful to investor's returns.

Today, I'll talk about one such example. But instead of showing you an example of a dividend cutter I will show you how a dividend raiser has hurt investors by keeping its payouts artificially low.

Consider CenturyTel (CTL), a telecommunications company and broadband provider. It's a dividend aristocrat, having raised its dividend consecutively for at least the last 25 years. It has raised its dividend by exactly a penny every year since 1999. That equates to about a 4-5% raise per year. Last year, CTL raised its dividends by 519%, including a one-time special dividend, and is on track to deliver a 73% raise this year (without the special dividend, this year's increase would be a whopping 977%).

Sounds impressive, right?

Only until we look at how fast earnings per share (EPS) was growing and how much of it was being paid out yearly. EPS grew at an annualized 6% pace. So far, seems consistent with the 4-5% growth in dividends. However, the payout rate has averaged 10.1% for all years from 1999 until the big dividend raise of 2008. The maximum payout rate during this period stayed below 16%.

This is a rip-off of shareholders money, by my account. Especially considering that during this period, the share price has grown from a low of $35.19 in 1999 to a high of $42 in 2008 -- an annualized 1.9% at best.

While I'm encouraged to see the payout ratio grow to something more significant, I recommend investors investigate the reasons for the sudden change in behavior. Was it due to a policy change? Debt covenants that have since expired? New management? Lack of new investment opportunities -- does the company consider itself mature now?


In general, it's easier to put a price tag on companies with a steadily growing flow of dividends and a stable payout ratio. However, a very low payout ratio needs to be investigated and discounted appropriately, since it's a possible indication that management might not have its dividend policy aligned with that of a serious dividend investor. At the very least, a company that withholds earnings from shareholders needs to have a credible growth plan and a history of shareholder friendliness.

Disclosures: No shares in the above mentioned security at the time of writing.


When to Sell?

"Should I sell?" A fellow investor recently asked me this question. It is probably the hardest question for a long-term, buy-and-hold investor to answer. His question was about bonds in particular. He had some nice gains and was tempted to sell, since his bonds had stopped moving up for a while.

I thought for a bit and told him to hold on. But I didn't do a great job at explaining why.

After thinking some more, I came up with the three rules below. They may sound familiar. But bear with me because I arrived at this answer by distilling a few things I've experienced, learned, and believe in; not by repeating what others have written. If my rules sound similar, well, I won't get credit for being the first one to write about them.

So, here's a better answer to "when to sell". Sell when:
  1. The price has surpassed intrinsic value.
  2. Fundamentals have changed.
  3. When you are bored in a rallying market.
Price surpassed intrinsic value. As an investor, you should have calculated an intrinsic value (actually, a range of values) that you estimate a stock or a bond is worth. When market prices get ahead of them, sell. This is easier to do for bonds and other fixed-income assets than stocks, simply because for these you know precisely what the intrinsic value is. So, for a bond, if the price is above par plus future interests discounted appropriately (with a fudge factor to account for the probability of you never getting them), sell.

I usually sell bonds or bond funds when price is above par and I have doubts about the fundamentals (see also item 2. below). I never buy bonds at par, I always insist on a discount, no matter what the "fair market value" might be. But this is just me. Therefore, to remain consistent, I begin considering selling my bonds when they reach or go above par. If I'm positive of their fundamentals and the yield is still attractive (which, at an above par price they seldom are), then I may hold on for a little bit longer. But if the price gets ahead of par plus future coupons discounted, then the bond is gone.

For stocks, this rule is hard to apply. I usually start with rule 2. below. But if prices get too bubbly, I'll will probably sell at least my weaker companies.

Fundamentals have changed. If my estimates of the prospects of a company change, I usually reassess the value of my holdings. If the new intrinsic value is below the current price, then rule number 1. above dictates I should sell. It's that simple.

Reasons why fundamentals change abound. A few examples: critical management leaves on bad terms; the annual or quarterly report becomes hard to decipher; fraud is found to be pervasive within management; the assumptions about future prospects change materially.

One reason I usually don't worry much is if a product or line of products fails to materialize. That's because I mostly ignore promises of future products in my analyzes. So a new drug not being approved by the FDA should not have a material effect on my analysis of a pharmaceutical company, even though it typically does for Wall Street analysts. The same is true of new "hot" products such as Apple's next iPhone or whatever else tech companies might have promised investors.

When you are bored in a rallying market. I know this is controversial, but let's be honest, everyone gets overconfident sometimes. Saying you shouldn't time the market is good advice, but it's almost impossible to follow. So let's make sure that if you do time the market, that at least you don't attempt to pick bottoms or tops, but that you simply have a profit and a reason to sell.

Originally, I was going to write this rule as "when you need to raise cash in a rallying market". But what is needing to raise cash if not being bored with the returns on your portfolio? Sure, emergencies. But that aside, raising cash usually means you found other investments that may be more attractive.

So, it's fine to rotate out of some investments if the market is rallying, you're ahead and you've found something better. The rule is: a) the market is going up and b) you have gains in the securities you want to sell and c) you have a more attractive investment to make and d) you follow the rules for buying (more on that later).

That way, you avoid: a) selling at major bottoms, b) losing money, c) limiting your gains, and d) buying without a reason.

And that's the only time you're allowed to be so foolish as to try to time the market or sell for a bad reason (being bored).

But let me re-iterate the third rule again: when you're bored and trying to time the market at least make sure the market is in your favor -- sell high -- sell during market rallies at a gain, never at a loss. Then, make sure you have something else in mind. Cash is not an option. Taking your gains just because you have them doesn't make sense. After all, if you did your homework right, you should expect to have even larger gains in the future. So sell only because you've found something better. And again, at a profit.

I hope my fellow investor didn't sell his bonds, but that if he did, he sold them at a gain to immediately buy something better, perhaps an undervalued stock of a profitable and growing company.


The Myth of Dividend Payout Ratio Part I: Is Lower Necessarily Better?

Everywhere I look for dividend investing commentary I always read the same "absolute truth": a lower dividend payout ratio is better. It's almost a dogma among dividend investors.

Intuitively, it makes sense.

The payout ratio is the portion of earnings that a company pays out as dividends. So, if you expect your dividends to be paid consistently and to not be cut or shrink over time, a low payout ratio makes sense: there's a margin of safety; if the company earns less one year, it will still be able to pay its dividend if the payout ratio is low.

The idea of "low payout is better" then is that companies can smooth their dividends if they retain cash for a rainy day and allow ample room for earning shrinkage by paying out less of their earnings in dividends.

So what's wrong then?

Well, two things. One is a principle thing and the other one is a historical fact:
  • In principle, all money not needed to be reinvested in the business (to keep the business afloat, such as replacing inventory, fixing up plant and equipment, etc) belongs to shareholders. So, management has no moral right to keep more than they should of shareholders money.
This can be debatable, especially if each dollar retained by the company gets translated to an extra dollar of appreciation of their share price.

As for the the historical fact:
  • Historically, smoothing out earnings hasn't provided any evidence that it can prevent or reduce the likelihood of dividend cuts. Consistent dividend growth and reliable payouts are the exception, not the rule.
At least that's what Ben Graham's investment company found out after many years of market research. The results are in Graham's book, Security Analysis. Consider this passage:
The typical investor would most certainly prefer to have his dividend today and let tomorrow take care of itself. No instances are on record in which the withholding of dividends for the sake of future profits has been hailed with such enthusiasm as to advance the price of the stock. The direct opposite has invariably been true. Given two companies in the same general position and with the same earning power, the one paying the larger dividend will always sell at the higher price.
(Security Analysis, 4th edition, Benjamin Graham and David Dodd, 1940)
The authors go on to explain that the theory of lower payout ratio may sound good, but that it fails to provide adequate protection regarding dividend cuts, and that therefore dividends should just be paid out as much as possible and it's the investor who should worry about smoothing his earnings over time.

This smoothing can be achieved by investors by them not relying on a fixed and increasing income stream from a dividend portfolio but instead relying on a fraction of it and saving the rest for a rainy day.

If the investor is not depending on the income stream for a living, then it's even better to get the maximum dividend now instead of later since an intelligent investor will have an easier time allocating the income proceeds into the most interesting investment opportunities at any given time than would individual companies.

In the book, Graham shows a few examples of companies that attempted to smooth out their dividends just to cut them some years later.

In another post, I will go over a current example where the smoothing out of dividends was detrimental to shareholder's returns.


Free Market Commentary Usually as Good as Its Price

Don't believe everything you read out there. Errors, omissions and mistakes abound. Here's is this week's example. Consider this news article by Zacks, "Brazilian Manufacturers Slightly Relieved".

The news piece claims that some Brazilian companies should benefit from consumer spending, "These include home appliances retailer Companhia Brasileira de Distribuicao (CBD), [...] cement maker CEMEX S.A. de C.V. (CX) [...] and state-owned steel producer Companhia Siderurgica Nacional (SID) ".

Wait a second.

CBD is not a "home appliances retailer". It's a supermarket chain.

CEMEX is not a Brazilian company either. It's Mexican.

SID is not state-owned and hasn't been since 1994.

And this comes from a "respectable" market and security research company.

Is this even relevant to the article's point? Probably not. But these errors are so simple and straightforward to catch that they show no editorial quality control whatsoever.

If this were the only example where errors, bad data or bad advice happen on the Internet (or on TV or other media in general) we'd probably be okay. But beware, it's out there. This one was only too obvious not to comment.


A Good Company in a Poor Industry

Even though management is crucial for the success of a company, the type of industry and business it is in is also very important. Warren Buffett once said that he prefers to buy companies that are so easy to run that they can be operated by a fool, "because someday a fool will". He also said that "when a management with a reputation for brilliance tackles a business with a reputation for bad economics, it is the reputation of the business that remains intact".

Why am I saying this? Well, because I think I've found a great company with great management but in an industry with relatively poor economics. And I'm not talking about airlines.

Consider SYSCO (SYY), a food distributor company. SYSCO sells prepared and raw food and food-related products to restaurants, healthcare and educational facilities, lodging establishments and other food service customers.

SYSCO's results over years have been pretty good. In the last 10 years SYSCO has:
  • Increased its dividend by an annualized 17%.
  • Increased earnings-per-share by an annualized 13%.
  • Reduced outstanding shares to 600 million from 670 million.
  • Had an average ROE of 31.9% (with an average leverage of 2.91).
  • Grown sales an annualized 8%.
  • Grown free cash flow by an annualized 14.6%.
A pretty impressive achievement for a company of its size, $12 billion market cap.

On the other hand, consider the difficulty of achieving these results. This company depends heavily on commodity prices (fuel and food), has limited pricing power, achieves only minimal differentiation from competitors based on services since its products are very similar to those of competitors, and is highly dependent on the North American market for the bulk of its earnings.

Food Prices and Pricing Power.
SYSCO buys food to sell to customers. Therefore it's exposed to food prices. In its most recent 10Q report, they say "Sysco attempts to pass increased costs to its customers; however, because of contractual and competitive reasons, we are not able to pass along all of the product cost increases immediately".

Considering that their products are similar to that of the competition, they must compete on higher-quality services and price. Therefore, the economics of this type of industry are not great. Compare that with products such as J&J's Band-aid, P&G's Gillette razors, WD-40 and Coca-cola. No one will choose a cheaper brand, even if the difference in price is 10, 20 or 30%. But customers of SYSCO do care if prices are 5 or 10% lower, especially if the product is very similar.

Fuel Costs. SYSCO is also exposed to fuel costs, since they deliver the food using their own trucks. SYSCO attempts to hedge fuel costs by entering forward diesel contracts. About 70% of their fuel expenses are on the basis of fixed-price agreements. However, this means they need to make bets on the direction of oil prices. In 2008, the company entered forward-contracts when oil prices were high and thus had to pay higher prices for diesel than spot market prices during the year. Management says
We periodically enter into forward purchase commitments for a portion of our projected monthly diesel fuel requirements to lessen the volatility of our fuel costs due to changes in the price of diesel. In the first 39 weeks and third quarter of fiscal 2009, our forward purchase commitments resulted in an estimated $50,000,000 and $22,000,000, respectively, of additional fuel costs as the fixed price contracts were higher than market prices for the contracted volumes.
(emphasis mine).

On the flip side, now that oil has dropped form last year's peak, SYSCO is enjoying lower fuel costs and is thus immune to increases for the duration of the current contracts. Over the long run, I expect such agreements to have no positive effect on earnings, as the ups and downs in the price of the contracts serve only to smooth volatility in fuel costs, but does not reduce fuel costs (to understand why, just think of the investor on the other end of these contracts).

Hence, SYSCO is squeezed between higher food prices that can't be passed on to customers automatically and higher fuel costs that are hard and expensive to manage. This means an investment in SYSCO is probably a bad hedge against inflation.

In fact, management recognizes this much: "Prolonged periods of high inflation, such as those we have recently experienced, have a negative impact on our customers, as high food costs and fuel costs can reduce consumer spending in the food-prepared-away-from home market".

SYSCO's net earnings as a percentage of sales has recently been in the 2.6 to 2.8% range. Compare that with Coca-cola's 18% and J&J's 23%.

Conclusion. SYSCO is a well-run company, in a stable and somewhat profitable market with large and stable demand. The company has a 16% share of a $231 billion a year market. However, the economics of this industry are not appealing and a fool could not run this company successfully for very long. Therefore, investing in SYSCO is making a bet in its management and in a deflationary to mild-inflationary times ahead.

At reasonable prices, SYSCO is an appealing buy. But it's a company one needs to watch the fundamentals very closely and be ready to sell when fundamentals deteriorate.

Disclosures: I own SYSCO at the time of writing.