Entering and exiting stocks and portfolio sizing cntd.

In an earlier post I shared my thoughts on portfolio sizing.

I have backtested a simple strategy that scales exposure to Berkshire based on P/BV - with full exposure at 1.2 BV (buyback limit) and no exposure at 2.0 BV. The latter is the best proxy for intrinsic value that I can think of that does not benefit from hindsight.

A not-very-precise backtesting of this strategy with 100 shares (1 lot) as a step yielded slightly better results than having constant full exposure to BRK. The average exposure was slightly above 50%, thus interest income from cash or investing on small margin could increase results.

For harder-to-value stocks, I will:

• estimate lowest possible price and plan full exposure at such price
• target zero exposure when price reaches intrinsic value estimate

I have been setting up limit orders on both sides of my current positions according to this system for some time, and it seems to work reasonably well. It captured recent dips and spikes on VRX and GILD, and a smaller one on BRK but is still recovering on KORS. I am still thinking about exposure during earnings announcements (esp. after hours). And also I am deeply thinking about Matt Levine's remark that having limit orders laying around is akin to giving away free options.


For couple of reasons I plan to spend much less time researching and tracking my investments from now on. The first one is that I am increasingly wary about overall valuation of US stock market and USD. As (hopefully) stock prices catch up with my intrinsic value estimate, my sell orders would execute. In last few weeks, my leverage on IB account already went down by ~20% thanks to that.
The second reason is that I came up with a small project that I want to spend more time on going forward.

I plan therefore to periodically update my valuations, post portfolios and reenter known stocks when prices drop, but do not research new companies and spend less time tracking developments in existing investments. I will perhaps also shift away from USD, possibly into some european or EM ETFs, but I am not yet sure about their valuations. I may even consider US fixed income given higher rates, but probably with short maturities.

Gilead 2016 results

While Gilead posted surprisingly good results for 2016, the very low 2017 guidance sent shares down after hours and on the following day. My very simple DCF model, assuming flat HIV and 30% annual decline of HCV for the next 7 years still yields $120 per share value with, I believe, conservative assumptions. Assuming no R&D spending and severe decline in each line of business (15% HIV and 40% HCV) the shares would still be worth $44, compared to $65.6 price today. I still see more potential for reward than downside risk in Gilead, at current share price.

The poor guidance (perhaps too conservative?) lowered my expectations of firm value, but it is still above current share price. Contrary to many others I do not blame management for executing share buyback for prices higher than current price. But what confuses me a bit is the fact that they are now deemphasizing buybacks in the earnings call. I would rather hope that they double down on them, while still leaving room for sensible M&A and R&D. But I can live with shareholder returns in the form of dividends, and I expect that positive news (M&A, slower decline) will improve sentiment in the stock.

In the meantime, I meaningfully added to my position, with ~$70 average cost.

Tesla catalysts

The Tesla saga is an interesting one, and I continue to track it closely.

Here is a nice write-up from Whitney Tilson. I think it nicely sums up short thesis (mostly written down by his friends, actually) as well as why it is risky to short Tesla. I do not like the the potential for unlimited losses from straight shorting, therefore I decided to short Tesla using Jan'18 and Jan'19 puts.

Besides solid short thesis I think it is equally important to identify catalysts. While I think that Model 3, driver assist functions in 2nd gen hardware, let alone autonomous capabilities will be significantly delayed, I don't see these events by themselves necessarily impacting Tesla stock price. I see following catalysts though:

1. While the date of the Tesla's annual report was still not given (perhaps a sign on its own), it should be out in February. With consolidation of Solar City, it will not look well, neither from income, nor balance sheet perspective.
2. The ZEV credits will be increasingly hard to earn, with Chevy Bolt and subsequent 'real' electric cars coming to market.
3. While Tesla fans await Model 3, the demand for Model S/X will likely dry, and we may see declining Y/Y revenue numbers. Losses will likely increase too. It will be tough for Tesla to say that Model 3 is on track and successfully upsell to Model S/X at the same time.
4. The 2016 DMV CA disengagement report suggests that Waymo is way ahead in self driving capabilities, with GM Cruise being a second one. Based on Waymo's October 'production' plans and GM's partnership with Lyft, I see one or both of them deploying a fleet of self driving cars to a ride-hailing program, similar to Uber, this year. While they will likely still have human safety drivers, I expect it will have clear path to driverless operation and thus it will likely be priced competitively to Uber/taxis. When this happens, the illusion of Tesla's leadership in self-driving capabilities should be clear.