your mileage may vary...

“Not only may it be impossible to predict the future success of games based on the first months of sales. But the quality of upgrades may ultimately turn out to have a decisive role in determining long-term success of a title. This makes valuing a mobile app franchise a very tricky business indeed.” —

I find it laughable that anyone would think the predictability of success in this industry is predestined. Apps have made software disposable. It’s the new paradigm. When you pay $0.99 for something, you have no problem throwing it away the minute it loses it’s appeal. The only lock-in you have is to the platform.

May 18

Kevin Smithen, an analyst at Macquarie Capital, an investment firm in New York, estimated the $12.5 billion purchase price represented a $4.5 billion value for Motorola Mobility’s patent portfolio, $3.2 billion in cash the company holds, a $3 billion handset and TV set-top business, and $1.7 billion in net operating loss tax benefits it has been unable to use.

Willens estimated the $12.5 billion deal will include $3 billion in goodwill, or the value Google expects to generate from Motorola Mobility’s brand, know-how and other intangibles, not including the patents. He added that Google would be able to immediately use the $1 billion in U.S. operating losses absorbed from Motorola Mobility, thanks in part to how goodwill is amortized.

May 11

“I got the call 2 seconds ago. There is a ship’s horn when you answer. Then a guy excitedly starts telling you that you have won 2 boarding passes…That is as far as I got. I hung up.” —

May 09

“My tie was the symbol of how serious and important a year this was, and I wore it every day to show this.” —

So, by extension, Zuckerberg believed that meeting with analysts wasn’t serious or important. That’s a data point you should consider when deciding to invest in the Facebook IPO.

As I’ve said a few times on the podcast, you want a management that is willing to tell Wall Street to go screw themselves when they think the Street is being short-sighted on how value is being created. You also want management who is willing to intelligently buck convention.

But I’m a little concerned that Zuckerberg doesn’t think that IPO’ing the company is a big deal. Because it is. From here on out, Facebook is going to have to hit quarterly targets.

Those stock options that have been handed out to all the employees are going to be priced in real-time every day and will have a direct psychological impact on his employees. And institutional investors are going to be badgering the company with questions. “How’s the quarter going?”, “Why did capex spike this year?”, “What new products are in the pipeline?” “Why is average ad revenue per user down one pico-penny?”

And the truth is, there’s going to be a day where you want those institutional investors on your side.

I don’t think wearing a hoodie to a meeting is a sign of Zuckerberg’s ability to create value. And I don’t think I would’ve been offended if he showed up to my meeting in a hoodie. But I do think this tempest in a teapot is a sign that he’s not in Silicon Valley anymore.

Apr 30

Pinboarded, archived and included on the hotlist of things that I consult on a regular basis. After studying logical fallacies and behavioral biases in one form or another for almost a decade, I’m still amazed at how often I find myself falling victim to them.

Apr 13

Jamie Dimon’s Letter to Shareholders’ is on the “Must Read” list

Every time I read one of Dimon’s letters I come away with the impression that he is one of the rare few who truly understand what leadership means.

Apr 02

Computer Hardware Industry Overview

I had to put together some charts for a client a few weeks ago that I thought some people might find instructive. Basically, I was answering the age old question, “why is this time not different?” It seems whenever people miss out on big gains, their initial instinct is to ask, “how can I get the big gains next time?” instead of “are these gains sustainable?”

The gains, of course, are in Apple. We don’t own it, and the client was questioning our logic. And you only have to look at the first chart to understand how truly extraordinary those gains are. Here, we compare the Enterprise Value (Market Value of all outstanding shares + Total Debt) of Apple to the other peers in the Computer & Peripherals industry of the Value Line Survey.

Note that all data in these charts are reflective of Value Line numbers and are as of 12/31/2011. All plots generated using R and ggplot2. Click on the pictures for larger resolution.

You can see that once you get past Apple and IBM, the valuations fall off quite dramatically in this space. These two companies have definitely been the winners of the last decade in terms of capturing the attention of Wall Street. But how have these lofty valuations been achieved?

In the next graph, we see the relative value that Wall Street is assigning to Sales growth. Looking at the left hand picture you can see what Price-To-Earnings multiple is being assigned relative to the previous five year’s sales growth. It becomes apparent that despite the incredible sales growth over the last 5 years (including weathering the financial crisis) most of the companies’ multiples appear to have no correlation with sales growth. This means either:

Looking at the Price-To-Sales ratio (how much you can buy a share for relative to each dollar of revenue the company makes) gives a more rational picture. Still, no one would argue that the relationship between Sales growth and Price-To-Sales is a strong correlation in this industry. It’s just stronger than the other stuff you can look at.

If we compare the P/E and P/S ratios to the Earnings growth over 5 yrs of these companies we get the same picture. Lots of clustering towards a median growth rate of around 10% while valuations are all over the map. Apple has actually crushed the competition but doesn’t garner much respect relative to its competitors on a multiple basis. Each dollar of Brocade’s earnings seem to be worth more than Apple’s earnings dollar. This is probably because it’s hard for a $400B company to sustain those kinds of growth rates unless you believe that every citizen on earth will soon be eating iPads for three meals a day.

Finally, we look at the Returns on Equity for the industry. This measures the return on shareholder investment (book value) in the company. Think of it as the equivalent to a bond yield if you had bought the company for the book value of its assets less its liabilities. Apple is no longer the leader of the pack, IBM and Dell far outpace it since Apple has chosen to keep so much cash on its balance sheet. You can see now why investors were pushing Apple to issue a dividend. What’s more interesting is that the majority of these companies weren’t generating a solid return on capital over the last 5 years. Half of them earned less than 10%. IBM has probably done the best job in the industry of generating solid growth over the period while returning capital to shareholders. Dell has been very good about returning money to shareholders (via buybacks), but it’s less impressive in generating solid growth.

The main takeaway from all these charts is that the range of multiples in the industry are not based on fundamentals, but are a function of investor sentiment about the future. The disparity of multiples in the space are a testament to the fact that the past will not resemble the future.

So, what’s an investor to do? Spend most of your time looking at the top line sales and trying to assess the sustainability and growth profile of that going forward. How competitive are the products? How are the competitors responding? What will the landscape look like 5 years from now? Once you get that, you can take a stab at forecasting margins with the understanding that the error bars are pretty high for forecasting in this space.

There are values to be had in this space. They just involve coming to grips with more operational risk than in other sectors. The payoffs in the tech space are much shorter-lived and much more assymetric than most investors give it credit for.

Mar 15

R Code to deal with Dates in datasets

I deal with a lot of small timeseries datasets. I like to use R as much as possible to deal with these, since I can then quickly reuse the scripts when the data updates. One of the more common, and mundane problems I run into are date formats. The as.Date( ) function does a lot of heavy lifting with date/time data, but one of the more common date formats that it struggles with is “MMM-YYYY” where MMM is the three letter abbreviation of the month (e.g. JAN for January). This is an issue because as.Date( ) expects a day somewhere in the data. The function below inserts an 01 for the day in the middle of the MMM-YYYY date string. So, a string like “JAN-1999” would be transformed into “JAN-01-1999” and be ready for date conversion using as.Date( ).

It’s not exactly clever R code, but since I’ve used it at least 10 times already, I figured I’d post it here in case anyone else found it useful.

## function to prep Date series in format %b-%Y for
## Date conversion by inserting a '01' for the day
## expects a vector of character dates. Note that
## separator character is irrelevant.
insertDayDate <- function(dateseries) {
  if(class(dateseries) != "character") {
    return(paste("expected character, but got ", class(dateseries)))
  }
 
  regexdate <- function(x) sub("([a-z,A-Z]{3})(.)([0-9]{4})","\\1\\201\\2\\3",x)
  datemod <- sapply(dateseries, regexdate)
  return(datemod)
}