Trading diary 20 December 2017
The experiment has started out with heavy losses. The net liquidation value of the experiment account is currently $81,000, down 10%. Two reasons: shorting crypto-related stocks as well as a large position in Criteo. I am inclined to reduce the overall risk level of the portfolio to contain losses.
The crypto bet
Most indicators I look at indicate a peak in attention to Bitcoin and other cryptocurrencies around 7 December. This includes # of Twitter mentions, Google Search query counts and autocorrelation models. I’m also noticing a small shift in the mindset of crypto enthusiasts – there is a certain lack of euphoria.
However, crypto prices have stayed flat (for Bitcoin) or continued to race ahead (for most alt coins). Possible explanations include a launch of futures trading on CBOE/CME, selling restrictions, Asian buying and a gradual switch to institutional buying.
MGTI has almost doubled in the last week due to excitement about its new Nordic crypto mining hub, causing losses of about 5% of the portfolio. XNET and SRAX have stayed more or less flat. I will continue to reduce my positions as losses mount, and vice versa. I continue to think these stocks are worth a fraction of what they are trading at.
Criteo is down roughly 30% from just a week ago due to a weaker-than-expected guidance from Criteo as Apple has blocked the work-around that Criteo previously relied on (HSTS).
The guidance was weaker than my worst-case scenario. ~37% of Criteo’s revenue comes from non-app mobile and Safari has a mobile market share of ~25%, adding up to 10%. How did they get to their guidance of 22% impact to revenues? Possibly because of higher conversion rates for Apple users. Or possibly because of management wanted to guide low and beat expectations.
What I think will happen
There will be attempts to work around Apple’s ITP, such as:
- Criteo working with publishers to store user information in first-party cookies
- Page redirects so that Criteo cookies become first-party cookies
- Publishers forcing users to turn off ITP to visit their websites
These methods will either work, or they will not. In a worst-case scenario, 22% of revenue is lost. Given a relative lack of operating leverage and underlying revenue growth rate of 30% it is possible that 2018 EPS growth will be zero or even lower.
Management is guiding for zero additional impact of the European GDPR regulation in May 2018. The industry has prepared for GDPR for two years, and in practice it will just lead to a slight change in how users opt in or opt out of user tracking. The impact should be minor. More concerning would be an implementation of the so called “ePrivacy Directive”, which would require user consent for sharing user data. But whether the ePrivacy Directive ultimately turns into a law is highly uncertain.
The bottom line is this in my view: Do users want ads to be relevant? Do publishers want ads to be relevant? Do clients want ads to be relevant? The answer should be a resounding yes. I could care less whether cookies are being collected if it takes away unwanted clutter. Users want free services, and for free services you need ads. The link between advertising and sales has never been clear in any medium. If anything, Criteo’s performance-based ads are the best that has ever happened to the advertising industry. Now clients are able to track whether ads are effective or not, through the help of third party verification. In an environment where clients are afraid of ad fraud, advertising dollars should be moving towards the performance-based ads that Criteo offers and away from traditional media.
What attracts me about Criteo is a strong franchise (even competitors call their technology superior) in a market that keeps growing in excess of 20% per year. By end of year 2018, Criteo will have a prospective growth rate of 20-30% against a forward EV/EBIT of <7x (assuming 6% operating margin). If I am right, a double is near-certainty. Given that ITP is largely a one-off event that will be worked through by 2Q18 – I can’t see why not. Acquisitions in the industry have typically happened > 2x sales, while Criteo now trades at less than 0.7x.
Criteo doesn’t have any debt and EUR 360 million in cash on its balance sheet. While the stock price fall is dramatic – it’s hard to go bankrupt when you don’t have any debt.
Much like Google Search engine has become superior by feeding it tons of data, the Criteo engine is superior thanks to its dominant market position. So network effects account for part of Criteo’s moat. Technology is a fast-moving industry, however, and things could change.
The fraud allegations are misplaced in my view. Criteo might cherry-pick ROI numbers in their marketing but clients are not stupid. The impact of individual ads can be measured through third-parties – and they do. Clients are generally happy with Criteo’s performance. Anonymous Analytics wrote an excellent rebuttal to Gotham City Research’s fraud allegations.
If the ePrivacy directive is implemented in the current format, it would certainly hurt the European business.
While Criteo’s technology is market leading and it has a niche product it is certainly reliant on the ecosystems provided by Google, Facebook and Apple. Criteo could also be impacted by regulation, notably in Europe. On other hand, with a healthy net cash position and at this share price, the risk-reward skew is highly tilted towards the upside in my view. Once the one-off effect of ITP is worked through by mid-next year, the stock will likely trade at a growth multiple again.
I have scaled down my bet on Standard Drilling. I believe extreme speculator positioning and technical exhaustion signals could lead to a pullback in the oil price in the short run. I love the company and the stock – just not seeing any near-term catalysts that could justify a 20% position.
I upped the stake in Criteo to 40% at an average price of EUR 22.5 to take advantage of attractive prices and what appears to be capitulation selling.
The short in Zhengtong was switched to Zhongsheng as borrow became available. I am seeing encouraging signs of both legs of the pair trade: sales of Japanese vehicles is slipping, while model H6 sales in November for Great Wall Motor was strong despite a phase-out of the old model. I might write another before year end explaining my thoughts on Chinese auto stocks and Great Wall Motor in particular.