J. Bryan Scott
Quantitative analysis of the drivers of Buffett’s investment performance
Frazzini, Kabiller, and Pedersen have released an updated draft of their quantitative analysis of the drivers of Buffett’s investment performance in Berkshire Hathaway over the past 30 years. Specifically, they quantify 3 factors that Buffett has long attributed to his own success: high quality stocks, reasonable prices, and leverage with impeccable terms.
We find that the secret to Buffett’s success is his preference for cheap, safe, high-quality stocks combined with his consistent use of leverage to magnify returns while surviving the inevitable large absolute and relative drawdowns this entails. Indeed, we find that stocks with the characteristics favored by Buffett have done well in general, that Buffett applies about 1.6-to-1 leverage financed partly using insurance float with a low financing rate, and that leveraging safe stocks can largely explain Buffett’s performance.
The paper is fantastic. Although it genericizes Buffett’s strategy, it also praises Buffett for having the foresight 50 years ago to start implementing the strategy and stick to it. These sobering graphs represent visually just how superb is Buffett’s performance:
Retail Gasoline Price Trends Over The Course of One Year
via Business Insider
The Dunning—Kruger effects of incompetence
Why is incompetence bad? Kruger and Dunning remind us that it’s not only about the poor output of one’s own work.
Kruger and Dunning proposed that, for a given skill, incompetent people will:
- tend to overestimate their own level of skill;
- fail to recognize genuine skill in others;
- fail to recognize the extremity of their inadequacy;
- recognize and acknowledge their own previous lack of skill, if they are exposed to training for that skill
I guess a lot of artists—great artists—their first album was their best album because that album was 26 years in the making.
— Irv Gotti, Producer for Jay-Z
Amazon offers a “No Rush” option for Prime subscribers
A great way for Amazon to reduce shipping costs.
Buffett on minimum tax for wealthy
Buffett clarifies his “minimum tax for the wealthy” philosophy: it applies only to those earning $1,000,000 per year or more. This is well above the $250,000 level that Democrats target.
Time-regressive capital gains tax
From Clayton Christensen:
We should instead make capital gains regressive over time, based upon how long the capital is invested in a company. Taxes on short-term investments should continue to be taxed at personal income rates. But the rate should be reduced the longer the investment is held — so that, for example, tax rates on investments held for five years might be zero — and rates on investments held for eight years might be negative.
Make your metrics easy to memorize
One of the most important responsibilities of an analytics team inside a startup is to evangelize key performance metrics to decision makers elsewhere in the company. This boils down to:
- Defining and focusing on a few great metrics
- Ensuring that decision makers understand roughly how these metrics are calculated and why they are the best metrics
- Ensuring that decision makers know the approximate value of these metrics at all times
I focus on #3 in this post.
Why is #3 so important? In startups, decision makers are constantly making decisions. Members from the analytics team usually won’t be there. Therefore, unless decision makers have a persistent idea of their performance metrics, they will inevitably make decisions that are data-ignorant. And a major goal of the analytics team is to avoid the company making data-ignorant decisions.
Phrased another way: When metrics—products of a good analytics team—are easier to use, customers—decision makers around the company—will use them more.
So what qualifies metrics as easy to use? You need to check off #1 and #2, but I won’t address those in this post.
A successful tactic I’ve seen is to make your metrics easy for decision makers to memorize. Once decision makers memorize metrics, they become more comfortable with them, use them more often, and apply them more correctly. This increases the leverage of the analytics team and makes it more effective.
Because your audience is not the type to memorize 68,000 digits of π, you have to simplify your metrics. But how?
For broad consumption, I like to communicate four pieces of information in a metric:
- Unit (e.g. users, dollars, days)
- Order of magnitude (e.g. thousands, millions, billions)
- Rounding to two significant digits
The decimal value of π: 3.1 (unit-less)
United States GDP: $15 trillion
World Population: 7.0 billion people
Expected odds of being killed in a plane crash in one year for an American: 1 in 11 million (unit-less)
I think it’s obvious why units and order of magnitude are important pieces. But why two significant digits?
- For memorization’s sake, fewer digits is better. Most everyone can recount that π is about 3.1 or 3.14 but few people can remember more. Worse, more digits can distract the reader from the highest order digits. People also tend to transpose digits of big numbers in their heads. So I’ll take the bare minimum number of digits, please.
- One digit is often not enough to do meaningful calculations for decisions that are within 10%. And for startups, almost all of your atomic metrics-based decisions are calls within 10%.
- Two digits is almost always sufficient for meaningful calculations.
- Three digits doesn’t add nearly as much incremental value, so we’ve already reached diminishing returns on a continuously decreasing function (average value per digit as a function of number of significant digits).
Workday prices IPO at 37x revenue multiple
Workday priced at a trailing revenue multiple of over 37x. Wow. Compare this to Facebook’s revenue multiple at IPO of ~20x.
Workday, a Pleasanton, Calif.-based provider of enterprise cloud applications for human resources and finance departments, raised $637 million in its IPO. The company priced 22.75 million Class A shares at $28 per share (original range was $21-$24, revised range $24-$26), for an initial market cap of approximately $4.49 billion.
It reports a $46 million net loss for the first six months of 2012 on $119.5 million in revenue, compared to a $36 million net loss on nearly $55 million in revenue for the year-earlier period. The company will trade on the NYSE under ticker symbol WDAY, while Morgan Stanley and Goldman Sachs served as lead underwriters.
WorkDay had raised around $175 million in VC funding from firms like Greylock Partners (11% pre-IPO stake), New Enterprise Associates (10.1%), T. Rowe Price, Morgan Stanley, IndoUS Venture Partners, Janus Capital and Bezos Expeditions. www.workday.com
—Term Sheet by Dan Primack
Google’s market cap surpasses Microsoft’s
This was bound to happen, but today it finally did:
Note that GOOG’s business is actually valued $23B higher (10% higher) than MSFT’s already, because MSFT has $23B more cash than GOOG, yet they have about the same $250B market cap.
MSFT: $64B GOOG: $41B
Better than MP3s
As an owner some moderate audiophile equipment, I’ve long argued that listeners make unknown sacrifices when listening to a recording an MP3. When using a high fidelity playback system, the difference between an MP3 and CD quality sound is easy to point out. It’s amazing the detail one can hear in a CD quality recording that is otherwise muffled by lossy audio compression. If you haven’t heard a track in CD quality, you literally don’t know what you’re missing.
There are other benefits too. MP3s accomplish the illusion of precision in part by adding background noise. CDs have no such background noise. With the clarity of a CD, I can achieve the same perceived volume from a CD recording at only 70% of the amplifier volume of an MP3 recording. This is healthier for the ears and makes for easier listening over long periods of time.
Very much to the credit of lossless compression algorithms, nearly all of these differences are imperceptible in casual listening environments with non-audiophile equipment.
Yesterday Neil Young was featured on David Letterman (via TechCrunch):
“My goal is to try to rescue the art form that I’ve been practicing for the past 50 years. We live in the digital age and, unfortunately, it’s degrading our music, not improving it … It’s not that digital is bad or inferior, it’s that the way it’s being used isn’t doing justice to the art. The MP3 only has 5 percent of the data present in the original recording. … The convenience of the digital age has forced people to choose between quality and convenience, but they shouldn’t have to make that choice.”
In concept I completely agree.
Young goes on to describe the ideal format to be 24 bit per sample / 192,000 samples per second. Xiph has an excellent write up of how for playback, this is excessive, even detrimental.
Conversely, the Red Book CD Standard was designed to precisely meet the limits of human audio detection.
Lossless compression became popular because it produced smaller file sizes, and smaller file sizes made all the difference over the internet. Even today with advances in lossless compression technology, my lossless files are about 5 to 6 times bigger than their 192K lossless equivalents. But what bugs me today is the lack of choice—it is nearly impossible to purchase and download a lossless audio file (Bandcamp is a notable exception). Ironically, physical CDs are about half the cost of an MP3 album despite their higher quality—a testament to the convenience of online distribution.
As internet speeds increase and the cost of data storage decreases, the incremental cost of providing CD quality sound will approach zero and we will inevitably see the return of lossless music downloads. But this could be a decade or more away—digital music is a notoriously low margin business and there are licensing issues, too. Increasing bandwidth costs by 5x is nothing to scoff at. It’s also not clear that record labels want to make CD quality downloads available.
To the music industry, this is a great opportunity for price discrimination. Audiophiles are notoriously (sometimes ridiculously) price insensitive. As an audiophile, I would pay a lot more for CD quality audio. How much? Probably 10x what I currently pay for lossy internet radio services.
While I disagree with Young on the level of quality required for audiophile listening, I do hope that more CD quality audio is available for download or streaming in the near future.
Presenting Warren Buffett’s “Gold Cube”
Recall from Warren Buffett’s 2011 letter to investors:
Today the world’s gold stock is about 170,000 metric tons. If all of this gold were melded together, it would form a cube of about 68 feet per side. (Picture it fitting comfortably within a baseball infield.) At $1,750 per ounce – gold’s price as I write this – its value would be $9.6 trillion….You can fondle the cube, but it will not respond.
Below is what said gold cube containing all the world’s gold would look like, with distinctions for the various types of gold currently in existence: