Thursday, June 28, 2007

Investing in Linux

You could not keep eager investors from bidding sky high prices for Linux companies back in 2000. I tried to explain that Linux was not likely to hurt Microsoft in the short run, and that something that is available for free is a tad difficult to turn a profit on.

Come 2007 and the promise of Linux to make a profit for investors looks much brighter. But the premier Linux stock, Red Hat (RHT), while recovering from $3 lows in 2001, is only worth 1/6th of its January 2000 high. Its main competitor, Novell, is also not getting high P/E ratios for its efforts. The Linux plays that are doing relatively well are seen with companies where Linux plays a supporting role, rather than being a revenue driver itself. Companies like IBM and Oracle.

Is Linux still suffering from fundamental defects in its profit model? Or are investors missing the boat at Red Hat and Novell? How should these stocks be priced?

Let's start with Red Hat, which is a pure open-source play with income primarilly from Linux. It just reported Q2 2007 results that disappointed some investors. [See my summary of the analyst conference.] Revenues were up 47% from year-earlier: sounds like a growth stock to me. Net income was $16.2 million if you like GAAP, and I do; they also claimed non-GAAP net income of $33.7 million.

So what is their market capitalization, end of day today? $4.3 billion. That is pretty hefty. Using GAAP, factoring out the Q2 income to $64.8 annual, you have a ratio of 66. If revenues are up again 47% a year from now, and if economies of scale make earnings grow faster than revenues (or GAAP EPS converge to non-GAAP EPS), the stock will look cheap in retrospect. But projecting rapid growth out more than a year or two in advance is a good way to get taken to the cleaners.

Linux has changed. It is still geek friendly, not general population friendly. But the general population is not running corporate server farms. Linux evolves faster than Microsoft Server or the IBM OS's. And you can charge enough for support and service and bundling features to make a profit. Linux can be had for free, but that works only if you are highly skilled. So corporations are willing to pay for getting a well-tested version and support services.

Oracle has proven, yet again, that it is smarter than we think. Oracle is more of a threat to Microsoft now than at any time in history. Why? Because instead of butting heads with Microsoft it went after weaker, higher-margin opponents. It is killing SAP and shoving IBM around too. Its database system is the jewel, but the strategy of creating a full service ecosystem around the database has given the company new life. Oracle's offering Linux is not so much a danger to Red Hat, or even Microsoft, as a demonstration of what the tech jungle may look like a few years from now. Oracle software runs well on Microsoft Server products. It runs well on Unix and Linux. It plays well with CRM and ERP. And if you are buying most of your stuff from Oracle, there is no need to get your Linux from Red Hat. For more dope on Oracle see my summary of their June 26, 2007 analyst conference.

So far the conversion of Unix to Linux is happening so fast that there is plenty of pie for everyone. Another player is Novell. They have a big pot of cash ($1.8 billion) and a stategy for providing Linux and other services, but have been hard pressed to show a profit on a regular basis. They generate more revenue than Red Hat, but much of that is from legacy products that are clearly in decline. If they find their footing it will probably be through acquisitions.

Another issue for investors is the swarm of small Linux players. There are lots of versions of Linux available, all of them free. Some probably have ambitions to become serious players, and nothing is in place to stop that from happening.

More data:

Red Hat investor relations page
My Red Hat page
Oracle investor relations page
My Oracle page
Novell investor relations page
My Novell page

Wednesday, June 27, 2007

Show Me The (Lost) Money!

Are American subprime mortgages the Black Swan that will lay low (for a while) our securities market and even the global economy?

Everyone can agree that if you buy a house and then sell it later for a higher price, you have made money in some sense. You might want to take out the transaction fees. You may want to take out real estate taxes and the cost of maintenance and upgrades. On income (as opposed to residential or purely speculative) property there is another whole set of calculations. But basically, you lose money if you sell for a lower price than you paid originally, and you make money if the sale price is higher.

In 2005 selling houses almost always resulted in making money. In 2007 selling houses is resulting in losses mainly for people who bought in 2006 or 2005, rarely for people who bought in 2004 or earlier.

Mental calculations of loss may be higher because people who are selling now are thinking about what they could have made if they had sold in late 2005 or early 2006.

So the reality is that not very much money, on a macroeconomic scale, is being lost in housing right now. Think about the scenario we read so much about: someone who took out a subprime mortgage, with no money down, in 2005 whose house has gone down in value and who has allowed the house to be repossessed.

How much did they lose? They invested no money down. They came out of the deal, two years later, with no money from the repossession of the house. They actually did not lose any money at all. In effect their mortgage payments were rents. We all tend to forget that while a house is mortgaged, we have a partner in the real estate, the holder of the mortgage.

In a slightly different scenario the house value may not even have gone down. Instead, after the ultra-low interest rate introductory period, the interest rate has jumped. Unable to make the new higher payments, the owners allow the houses to be repossessed. Again, the residents did not lose any money at all; they just had their hopes dashed. In this case the mortgage company has not even lost any money except to the extent that interest rates had been subsidized.

This whole scenario could change if we had a recession before the glut of houses for sale dries up. Home owners who sell could lose real money they put in as down payments and as principal payments on mortgages. In some specific geographies this has already happened: people and mortgage companies have lost real money. But there has always been a component of this in the housing and mortgage markets. It has been a long time since it has been a good idea to invest in residential housing in Detroit and other rust-belt cities.

So how did the Bear Stearns hedge funds lose so much money? They made a bet on mortgage default rates than seemed smart at the time. To the extent that there were potential real losses of money, they concentrated it in their funds. They bought collateralized debt obligations, CDO's, and other mortgage-releated securities.

How CDOs work has been covered by most major financial news organizations, so I won't go into that. I want to point out the contrary fact: On the whole, investors have not lost any money. Not yet. Not people who bought residential housing; not mortage companies; not holders of investments in mortgages.

What has been lost, in aggregate, is expected streams of income. Only those investors who bought mortgage-backed securities that concentrated the risks of subprime mortgages are suffering serious losses.

As a result, risks are being recalculated. Usually in situations like this, after risks have been under-assessed, they jump to being over-assessed. Which is safer for us all.

Monday, June 25, 2007

Housing Market Distortions

Most traditionalist economists believe that markets are always in equilibrium. I believe that markets are almost always out of equilibrium. The questions speculators have to answer correctly are: how much, and in which direction.

Comparing the housing market and the stock market can give a lot of insight into causes of disequilibrium. The stock market can be broken down into stocks that are heavily traded and stocks that are thinly traded. To a large extent the national housing market behaves like a heavily traded stock, but local markets trade more like thinly traded stocks.

Another useful dichotomy: is the trading auction style, or swap style? By swap style I mean the traditional markets as described by Adam Smith: a rational (or at least savvy) person at each end, trading something of value, when there is an elastic supply of the two items (one usually being money) to be traded.

Today there is an almost 10 month supply of unsold houses on the U.S. national market. Two years ago houses more often than not sold the day they were put on the market; houses under construction sold before construction began. Yet the economy is arguably stronger on the whole today that it was two years ago. Anyone who argues that the housing market, or sale prices of housing, is always at equilibrium is saying nothing. They are defining equilibrium to be whatever happens. Same for stock prices.

Suppose you are going to buy something; it could be a stock or a house. As long as the price is rising and continuing to rise it makes sense to buy as soon as possible. Maybe normally you would save up a 20% down to get a good interest rate on a mortgage; but with prices increasing, it makes more economic sense to put 5% down and pay a higher interest rate now, rather than waiting two (or ten) years to save up the down.

The same becomes true when prices are falling. Why buy house now, even if you want it and think it will be a good long term investment, if you think you can buy it for 5% or 10% less if you just wait 6 months?

Rising prices accelerate demand, which in turn makes prices rise more. That is one reason why stock prices and housing prices tend to be out of equilibrium. Falling prices tend to dampen immediate demand, which in turn causes prices to fall more. That is the other main reason prices are usually out of equilibrium. These tendencies are the basis of macroeconomic cycles and of stock market price fluctuations that may last hours, weeks or even years (or, with super-fast computerized program trades, fractions of a second).

Auction systems aggravate these trends, partly because of human psychology and partly because they create a short-term artificial scarcity or surplus. Price swings in thinly traded, illiquid stocks display this.

Buyers in the housing market aren't going to be in a hurry until prices start rising again. The exact turning point won't be obvious, partly because it will take place in different localities at different times. The willingness, and ability, of lenders to finance purchases may stall or accelerate the process of finding a bottom.

But there will be a bottom. Strangely, the longer it takes to reach it, and the deeper the retreat in prices is, the more people will think there is no bottom.

Wednesday, June 20, 2007

Buying Opportunity in Undervalued Microchip (MCHP)

Microchip announced yesterday that sequential revenue growth from Q1 2007 to Q2 2007 is likely to come in around 2%, disappointing investors who had relied on previous guidance for 5% growth. Earnings projections were not as bad. The press release said earnings per share are expected to be $0.36 (GAAP), which is only one penny short of prior guidance.

Investors, of course, had hoped Microchip would do better than guidance; it often does. So the stock plumetted yesterday (due to auction price movement amplification). I believe it is was already slightly undervalued before the news. Yesterday's news and price fall leaves it undervalued. (I own the stock but have no plans to buy or sell in the near future because it has appreciated as a percentage of my portfolio to the point where my rules do not allow me to buy more; but not to where my rules require me to sell it).

If you look at the Q1 results and analyst conference held on April 26, 2007 (See my summary), you'll see revenues were $258.2 million. A 2% increase will bring them to $263.3 million. In Q2 2006 revenues were $262.6 million, so revenues will be about flat year-over-year. GAAP earnings were $0.35 per share, so we can expect only a 3% year-over-year increase in earnings when Q2 2007 results are announced.

Can a stagnant semiconductor company be undervalued when its trailing price-to-earnings ratio is 23.37 (per NASDAQ, at a price of $37.70 per share)? I think there are three details that need to be considered.

First, at the end of Q1 Microchip had $1.3 billion in cash and equivalents. That is 15.8% of today's market capitalization of $8.2 billion. It is especially healthy when you consider MCHP pays out $1.12 per share per year in dividends.

Second, there is the reason revenues are falling below prior guidance. Demand in Europe has been weaker than expected this quarter. Memory prices have been low so management chose to pass up sales that would have hurt profit margins. Memory is an adjunct businesss for Microchip, which is best known for its microcontrollers and which has also moved aggressively into analog chips. Low memory prices are an industry-wide phenomena that has been widely reported.

Last quarter Microchip reported it had finally become the leading global supplier of 8-bit microcontrollers. But the fastest growth is in 16-bit microcontrollers. They report that this quarter they continue to have strong 16-bit microcontroller revenue growth. Hence the good net income expectations despite the dissappointing revenue growth.

Third, I believe that inventories across the industry are lean and that worldwide demand will be good in Q3 as electronic device manufacturers gear up for the holiday Q4 rush. So demand for analog chips and microcontrollers should be strong in Q3. In 2006 Q3 was in the opposite situation: end-users had overstocked in early 2006 and were trying to reduce inventories.

So Q3 year-over-year comparisons should look healthy and justify a higher price-to-earnings ratio than the stock has today.

On the other hand I don't see a period of hypergrowth ahead for Microchip. Steady growth with a very healthy dividend is the picture here. The microcontroller industry is very competitive, but Microchip has worked its way up through the ranks. I expect it to remain at the top; I like management's commitment to profitability.

More information:

Microchip Investor Relations Page
My Microchip (MCHP) Page

Monday, June 18, 2007

Akamai Acceleration?

One reason I don't own stock in Akamai Technologies (AKAM) is that the current price assumes very rapid growth; in other words, if you don't like to make assumptions like that, the stock is not cheap. But the company sports some serious business and technological brains. It is the sort of company that two years later, sometimes, makes you feel foolish for not buying it despite the high P/E ratio. Which stood today at 126 (per Nasdaq). In other words, earnings from the past year come to less than 1% of the value of the stock.

Akamai accelerates data exchanges over the Interet (see my Understanding Akamai). Most people who have paid for a broadband Internet connection expect quick downloads; if a page does not download nearly instantaneously, they are on to something else. For a site financed by ad revenue this is bad. For site that serves up such ads, it is very bad. For a site that hopes to sell people something, slow is a disaster.

According to Akamai, of the top 100 Internet retailers, over two-thirds use Akamai to accelerate their web shopping experiences. Akamai has quantified the value of this acceleration: an 11% increase in revenues after Akamai's acceleration technology has been installed. One customer, Motorcycle USA, saw an over 15% improvement in conversion rates (completed sales once a customer has reached a stage of interest). Revenues increased 30%. Best of all their call center volume dropped, allowing them to save money.

Watching my wife try to shop on the Internet is instructive. She browses, but about one-half the time ends up calling in her order rather than finishing it online. She can spend several minutes of employee time finishing an order. Akamai claims customers typically abandon a site after a 4 second wait. Paying Akamai to decrease the wait pays off.

This particular Akamai product is called Dynamic Site Accelerator. "The Akamai solution not only enables retailers to accelerate their dynamic transactions and interactive content, but has also increased browser to buyer conversion rates and reduces infrastructure costs by offloading traffic from customers' Web infrastructure."

For all that, Akamai's revenues were only $239 million in Q1 2007, with net income of $19 million. Even with a continuing ramp 2007, GAAP earnings are unlikely to top $105 million. At today's $7.9 billion in market capitalization, that gives a forward P/E ratio of 75. If the stock price stays the same for about a year and earnings do grow rapidly, today's price may seem reasonable in June 2008.

You might also want to view my summary of the April 25, 2007 analyst conference covering the results from Q1 2007. Management guided to $610 to $620 million in revenues. Management likes to project "normalized earnings," which is a lot higher than GAAP revenue. For instance in Q1 GAAP earnings were $19 million, but "normalized" earnings were $50.7 million. They are guiding to possibly $215 million normalized net income for the year 2007. If the GAAP to normalized ratio holds, GAAP full year net income will be only $70 million. A nice chunk of change, but not one that justifies the stock price.

So we have a rare current-day example of a great technology company that seems to be overvalued by investors. Akamai has great promise, but in my view it requires a couple of years of revenue and profit growth to justify today's stock price.

Thursday, June 7, 2007

Stocks: Time to Panic?

Whenever there is a selloff like we have seen this week people get worried. Seniors wonder if it is time to close their brokerage accounts and just live on CDs. Young people hesitate to put hard-earned money into stocks. Could it be 2000/2001 all over again? Could this be the equivalent of the lead up to the Big One, the 1929 free-market meltdown?

It is a good thing to think about investments before they are made. When you own an investment and it seems that others are willing to pay more for it than it is worth to you, it is good to sell.

But I see little risk in stock prices in general at this point. Few people (not even my wife) listened to me in 1997 when I said there something screwy about prices of Intenet stocks. Or in 2004 when I said housing prices cannot appreciate rapidly forever. So I am, so to speak, thinking aloud here, just clarifying my thoughts. I am much more heavily invested in California real estate than in stocks, but that was just luck, not prophecy or even planning. I bought when the market was low because I needed a home at a time that the market in my local area had hit bottom.

Today I bought Celgene (CELG) for the first time: it is a great company and the stock dropped a little bit down out of the stratesphere. With biotech companies there is always a great deal of risk involved, but I am happy to manage the kind of risk Celgene brings. (See my: 1. Celgene blog 2. Celgene analyst conference summaries page.)

Which DOW stock dropped the most, percentage-wise, today? Alcoa Aluminum, AA. It dropped
2.31% today. Is it an overpriced, substanceless high-flier? No, it has a price to earnings ratio (P/E) of 15.33. Which means trailing annual earnings are 6.5% of the stock price. Are there risks? Sure, the usual: a world-wide recession could hurt earnings, or energy costs rising faster than aluminum costs, for example. But there is an upside too. NASDAQ lists the forward P/E today as 12.25. That means the most likely, purely linear, scenario is that if you hold the stock for a year earnings will be 8.1 % of today's stock price. That is better than real estate is expected to do this year, and way better than owning a CD. Certainly safer than buying Celgene, but a bit boring.

Are there overpriced stocks? Of course, there are always overpriced and underpriced stocks, by whatever critia an individual my set. As I said Monday, I thought Dell was overpriced (See my Dell Stock Price Evaluation [June 4, 2007]). But not like Internet stocks in 2000.

I think reasonably well-informed investors can do better picking individual stocks than they would in index funds if they do their research and keep their heads. But I would have no problem recomending an index fund right now. Stocks on the whole are reasonably priced and likely to appreciate faster than U.S. real estate as a whole. Is there risk? Of course. It is good to assume that there is more risk than you think, as stock investors found out in 2001 and housing investors found out in 2006. But it is even more important to make sure that there is real value to what you buy, not just hype or an auction-fever driven pricing environment.

Monday, June 4, 2007

Dell Stock Price Valuation

In retrospect, of course, I wish I had bought some Dell stock back before the Internet age and then sold it at its peak in 2000. Even those investors who did buy at peak could have done worse: Dell's stock price was influenced by the Internet Bubble but was based on its proven ability to sell microcomputers. This year (trailing 52 weeks) the stock price has been all over, from a low of $18.95 (market capitalization: $ 43 billion) to a high of $27.89 (market capitalization: $63 billion). Today it is near the high end of the range.

What happened to the high price-to-earnings ratios of yesteryear? It would appear that Dell is no longer a growth stock. Its revenues, as given in its May 31, 2007 press release for Q1 Fiscal 2008, were $14.6 billion, up only 3% from $14.2 billion year-earlier. Dell claimed a 14% year-over-year improvement in average selling prices, so unit sales must have plunged in the vicinity of 11%.

There were some bright spots, with server revenue up 19% and storage revenue up 13%. Notebook computers did fine, with a 7% gain, but desktop revenues were down 6%. Since desktop revenues represented one-third of overall revenues, that decline balanced the other gains.

The company, of course, has a plan to do better. But none of it sounds all that great. Laying off employees may briefly help the bottom line, but it is not a growth indicator. HP is bound to try to press its recent advantage. Lenovo, Acer and other Asian price/quality leaders are eager to scoop up market share.

Notebook and desktop computers are a commodity business. Servers are pretty close to being commodities as well. Dell once defined how to run a commodity personal computer business, but now it is just one of the pack.

Smart investors should now be thinking about Dell as an old-school, low growth, integrated retailer and manufacturer. It should be paying dividends, but it is not. It should be honest with investors and the financial community, but instead it has stopped holding analyst conferences.

Given the risk, it should pay at least 2% over T-bills. So minimum 7% in earnings. That requires a P/E ratio of 14.29. Selling PCs is a seasonal business, so a full year of trailing earnings is the best measure, unless you are gifted with prophetic foresight on upcoming earnings.

Trailing earnings are listed by NASDAQ as $1.33 (See quote). So my version of a fair stock price is $19.00.

Optimists (and people who own the stock and don't like my opinion) have only two arguments for today's stock price. One is that the future is brighter than I see it. I admit to that possibility, but I think the downside probability needs to be taken into account as well.

The other argument is that you don't deserve a 7% return on investment. If you are willing to settle for less, sure, buy Dell at today's price.

Dell has not held an analyst conference since I started covering it, but you can see my summaries of other tech stock analyste conferences by clicking on this link.