Monday, December 26, 2016

How Green and Efficient is the Tesla?

             Much of the investment thesis behind Tesla, as well as a reason for the government subsidies, is based on the premise that its cars are more green and efficient.  Nonetheless virtually no one knows how much more green or efficient the cars actually are.  As the owner of a 2015 Tesla Model S 85D, I decided to find out by comparing my car to my son’s Prius.  Here are results of my experiments.
            Both my son and I drive similar distances daily around Pasadena, California (where Teslas are very popular).  Our trips are divided equally between freeways and surface streets.  He gets an average of 43 miles per gallon.  A gallon of gas contains the energy equivalent of 37.9 kilowatt hours (KWH) of electricity.  Therefore, my son’s Prius gets 1.14 miles per KWH.
            To compare, I charged my Tesla fully at which point it said I had a range of 238 miles and then ran it down to a range of about 30 miles several times.  The first piece of bad news for Tesla owners is that while my range dropped an average of 208 miles, the car only traveled an average of only 144 miles – less than 70% of the promised range.  To cover those 144 miles, I used 51.2 KWH of electricity which comes to 2.80 miles per KWH – more than 2.5 times better than the Prius.  An apparent smashing victory for the Tesla.  BUT, Edison, my electricity supplier, had to produce the power.  At the margin, Edison generates electricity by burning natural gas.  On average, only about 43% of the energy in the natural gas is converted to electricity, the rest being lost as heat.  The Prius has the same problem only more so.  For every gallon of gas burned only about 25% serves to move the car forward.  However, that fact is already accounted for in the miles per gallon number.  Therefore, to compare apples to apples, account must be taken of the energy loss at the power plant.  Doing so reduces the effective mileage for the Tesla to 1.21 miles per KWH - nearly identical to that of the Prius.
            From a green standpoint, the Tesla has another advantage.  To produce a KWH of energy from natural gas, Edison emits 190 grams of CO2.  To produce a KWH of energy from gasoline, the Prius emits 250 grams.  (Natural gas burns cleaner than gasoline.)  Consequently, the Tesla emits 154 grams per mile, compared to 220 grams for the Prius.  Furthermore, the Tesla does hold out the possibility of becoming greener as electricity generation moves to renewal sources like wind and solar.  On the other hand, it presents the problem of what to do with the used batteries.  Given the current scale of renewal generation and electric car use, these are both issues for the future.
            The bottom line is that from an efficiency standpoint the Tesla and the Prius ended up in pretty much of a dead heat.  To give the Tesla it due, it is larger and faster.  With respect to the green factor, the Tesla turned out to be 30% better in terms of CO2 emissions.  Whether that currently is enough to justify the Tesla’s snob appeal, its dramatically higher price, and its state and federal subsidies is another question.


Sunday, December 18, 2016

"Stationarity" and Investing: What Can We Learn from History?

            The best way to explain stationarity is with an example.  Suppose I want to estimate what the rainfall will be in Pasadena, California during the year 2020.  The annual rainfall can be thought of as a random process with a given, but unobservable, mean and random variance around that mean.  The process is stationary if that unobservable mean is constant.  In that case, the mean can be estimated by an historical average.  The longer the historical data period, the more accurately the mean is estimated.  But all that assumes stationarity.  What if the true mean is changing?  To answer that question, consider a related problem.  Suppose I want to estimate LeBron James’ scoring average during the 2020 NBA season.  Averaging his scoring average over this career thus far is likely to be an upward biased measure.  The true, unobservable mean for a player’s scoring average is not constant, particularly later in his career.  It declines with age and at some point, it goes to zero when the player retires.  The random process that generates the annual scoring average is not stationary.

            What does all this have to do with investing?  Much quantitative investing is based on compiling massive data records and then searching them for relationships that can be used to predict future returns.  Given the size of the data sets and the speed of modern computers, many significant historical correlations are invariably discovered.  But are they stationary?  The world is a highly nonstationary place.  Governments come and go.  Wars are won and lost.  Technology changes the way we live and work.  Economic policies are in constant political flux.  It is quite a stretch to think that in such a world historical correlations will be stationary.  In other words, the LeBron James example is probably more accurate than the rainfall example.  (Note even the rainfall process may not be stationary if there is climate change.)  If that is the case, it is hard to know which relationships that computers kick out can be trusted and which cannot.  In a non-stationary world, it is easy to exaggerate the benefits investors can derive from “big data analysis.”  It may simply uncover past relations that are no longer applicable.

            Warren Buffett once quipped, “All we have learned from history is that people don’t learn from history.”  But in a world characterized by significant non-stationarity, it is not clear what history has to teach.