Monthly Archives: July 2013

As an example of the negative selection process in the Obliquity methodology, let’s describe how I ended up excluding Tesla Motors.

a Tesla Model S electric car

Tesla Model S (credit: Tesla PR)

The company is very likeable. Electric cars are still an experiment, and not rationally competitive with petrol ones on a pure cost/efficiency basis, at least if you don’t count in environmental externalities. Even if you do count them, it won’t make the company’s success, unless it’s expressed purely in a tax credit form, because it’s unlikely anyone is buying a car above its rational price by computing the externalities in a rational manner at the individual level. People just don’t function that way.

That said, the concept of visible emission free — even if it may be power by a coal plant out of view — and general direction towards low environment impact seems to tickle many positive feelings, perhaps as an ersatz religion in the age where traditional monotheism is intellectually bankrupt.

Back to economics, all this means that “green” products can probably get away with up to say a 50% premium (for costly capital goods, it’s more for yoghurt) on the comparative “grey” technology. So I think Tesla has a product that can sell well with the sort of premium they’re targeting.

On top of that Elon Musk seems to have that rare Steve Jobs-style of talent of getting the full product, from concept to design and through production, with an uncanny ability to assemble a good team around him. So not only are Tesla cars managing to be green with a market-compatible premium (so the price is about right), they also seem to be very competitive design wise with existing petrol ones (that is a petrol Tesla would probably be well regarded too). This is no small achievement.

And he might even get the infrastructure challenge in the end (charging stations, battery replacement robots, or whatever). Although here I’m a bit puzzled by the “free forever” promise for fast charging stations. It’s a call to arbitrage (is it OK to load your Tesla there and use the energy to power your home for free?) and perpetual promises sound a bit dodgy. He might find a way to exit that quietly in due time.

Apart from that robotics and stuff is alluring, although the rest of the car industry is not that different here, and some of the Tesla spin might not be very substantial.

Anyway, i thought it was good enough for Obliquity London, and proceeded with the checklist, and that’s where it failed.

It failed unsurprisingly on “price action”, given the bubbly behaviour of the past year, despite a recent slight pullback. But that’s okay, it’s not our usual drink cans manufacturer. It is similarly failing “diversification”, being very clearly a single product (family) horse. Okay as well, but basically we must pass all other criteria to remain above 7 as per the portfolio rules.

Low volatility is also an obvious fail, though perhaps it could be talked away if we compare it to similarly speculative ventures (which I’ve not done in detail). That could also be mitigated through underweighting (expecting to reach the target weight through price increases) so cheating here may be okay.

Harder to say on “balance sheet”. On the face of it the debt is modest and manageable and as long as the share price stays high, Tesla has access to plentiful cheap funding by issuing new equity, which is another thing Elon Musk understands, having recently raised equity to repay a US government loan. That said the criteria includes the price paid for the said balance sheet, and here we’re in deep speculative territory and Tesla’s market cap is a vast multiple of its book value, so it either fails here or we have to value it another way. A back of the envelope valuation through comparison with other auto makers makes me think that Tesla is approximately valued on the basis it will sell around 1 million cars a year (we’re talking order of magnitude precision here). So it’s a good deal if ultimately it sells more than that. Will it? It’s thinkable, and definitely they’re not in the running to be a niche electric car manufacturer that keeps a sub-percent market share. Thus, it’s reasonable, but to make it investable they need to become bigger than that, without undue dilution on the way up, and to compensate for the chance of total failure, which includes surviving as a niche player, as that cannot support the current valuation. Another source of failure could be the second mover advantage, that existing or new players learn from their errors and capture their market. Although here I think Elon Musk’s talent may be unique enough, and I wouldn’t bet against him on that. So it’s close, but I think we have to say fail here.

Another surprising close criteria, possibly a fail, was “employees”. The reviews are mixed, most seem to love the startup ethos, tolerate the total commitment (no work/life balance), but we get a relatively bad vibe from the factory floor and an overall middling rating. The designers seems to treat workers and production supervisors as disposable dumb meat, which is not a good idea. They do make the product after all, even if they’re “just” doing robot maintenance and supervision — which they’re not, the production line is far from being fully automated. Also an adjacent theme seems a blanket rejection of the old auto industry culture, which is a pity. Questioning old habits is good, sometimes they do get out of date, but tradition often contains a lot of wisdom arrived to by trial and error, where people tried other ways, did the obvious errors, and where the consensus arrived at is evolutionarily close to optimal. The blanket rejection of the existing seems to imply they will make some costly avoidable errors. Unfortunate.

So overall, it’s a clear fail. It is definitely an interesting company but too speculative, and possibly too expensive, for the Obliquity portfolios.

Incidentally it’s got a high published short exposure, which seems crazy. Between the chances of actual success and of speculative spikes while you wait, it’s surprising that anyone wants to be short this. If it does fails, it will be years from now, long beyond the attention span of most short sellers.


Many people express worries about the exit for the Quantitative Easing (QE) process, but then can’t come up with a coherent narrative of how, short of gross incompetence, it could cause a problem.

The basics are very simple, if you’re a central banker you just follow the following algorithm:

  1. if inflation > target then sell some QE assets (this reduces core money supply, more directly than a rate change)
  2. else do nothing (keep them to maturity)
  3. repeat

The most likely scenario is that they do nothing given how hard it is to actually produce inflation outside of full employment. Letting things expire over many years is a very slow incremental process that is unlikely to cause any trouble.

Arithmetically it’s possible that #1 causes a notional loss for the central bank, as they may have to sell their assets for a lower price than they bought it if rates are higher at the time of the sale. But there’s two remedies to this problem (1) if rates go up on their own it puts downward pressure on inflation hence keeping it below target with no asset sales needed, (2) in the unlikely case where you’d get high rates and inflation, the balance sheet of a central bank is unconstrained so they can always either let it go negative, reset it, or book the loss to a perpetual intangible asset (call it “future growth” or what have you).

That works even if you believe QE had much effect in the first place, despite being a mere maturity swap of various kinds of money.