With the news of Elon Musk bidding for SolarCity (SCTY) on behalf of Tesla (TSLA) and calling the proposed deal a “no brainer” it interested us to see if that characterization bore scrutiny from a cointegration perspective.
Cointegration, of course, identifies time series linked by common factors – hopefully fundamentally economic ones where stocks are concerned.
Despite trying a range of techniques, at both intraday and end-of-day resolutions, there is no sign of such a relationship between these two firms. The reaction to the deal probably reflects this.
That said, could the stat be “wrong” somehow?
TSLA and SCTY are based on visionary but hitherto deeply unprofitable business models. TSLA sells at a sporty 27 times book (and that is after the poor market reaction to the deal) whilst even dowdy SCTY is approaching 3 times book.
For companies that have never turned a profit even at the operating level and can really only point to revenue growth (whilst liquidity ratios deteriorate) those numbers are acts of faith by investors. And the revenue growth is not exactly organic either spurred, as it has been, by a slew of tax incentives and subsidies.
In the case of SCTY the absence of ongoing funding of the type represented by this deal would be a death sentence. But Mr. Musk appears to have breathed life into it - initially at the expense of an army of short sellers but, by the end of the week, SCTY was pretty much back where it was pre-announcement. Brexit does not seem to have distracted the shorts a great deal.
So it is indeed conceivable that the time series contain so much idiosyncratic risk as to drown out the fundamental economic links (assuming they exist!)
And it’s also possible that Mr. Musk simply had to protect himself and Tesla from the reputational risk of a SolarCity failure.
Two issues already. Our thanks to those already subscribed. Concrete trade ideas, flash alerts and related content continue in the next issue on 4 July. We've had some requests to add Paypal to the payment methods - looking into it.
We suggest that in concert with the Premium Newsletter (€16/$18 per month) users get the pair discovery software tools and a running consultancy on deployment methods and strategies. That includes downloadable files covering selections that are 100% compatible with the software.
This combo of software + Premium Newsletter is also the most cost-efficient way for us to deliver a value-added service to users.
Anorak's corner: durability of cointegration
If cointegration is so durable why does it not last permanently?
The answer is that it breaks when the conditions that underwrote the relationship in the first place disappear. In simpler terms, those fundamental conditions changed (erm, “stuff” happens).
Cointegration is a linear association between multiple time series and most prone to break when shocks occurs. The timing of when some of these may occur is known (earnings announcements). But others are unheralded. Sometimes these shocks are temporary and the relationship later resumes; sometimes not.
So making generous assumptions of the length of time a cointegrated relationship will last is worse than watching the pair, and news flow around it, carefully.
However, for entertainment value the question “how long will it last” can be answered with a high degree of statistical confidence. This Google Book preview link to Professor John D. Barrow’s entertaining book “100 Essential Things You Didn’t Know You Didn’t Know” explains.
More proof that statistical confidence is not the only consideration to take account of with trading decisions.