In part 1 of this article, I set up my reasons for investigating Coinstar (CSTR) stock and set up what I see to be the key question that will determine Coinstar returns in the coming quarters; Whether Netflix's (NFLX) lost customers are opting for Redbox. In part 2, I looked at what state trends in Google searches might have to say about this question. In this third installment, I try to put it all together and reflect on how this analysis has informed my position with respect to Coinstar stock.
Putting it all together
If, like me, you invested in Coinstar thinking they were likely to benefit from Netflix's decline, I hope the results presented in parts 1 and 2 of this article have tempered your enthusiasm about that. Remember, competition between Redbox and Netflix is not necessarily a zero-sum game. Just because subscribers are leaving Netflix does not mean that they are using Redbox more. (I think the results so far presented should also make you question some assumptions if you are short Netflix because you think Redbox is the ready alternative, but note that I think there are other reasons to short Netflix.) The fates of the two companies' stock prices might serve to remind you that they are not necessarily in direct competition. Figure 7 shows the relationship between the stock prices since 2010. They have generally moved in tandem rather than opposite each other.
Figure 7. Daily opening prices for NFLX and CSTR, 2010-present.
Click to enlarge
If you have stuck with this article so far, I have one final interesting analysis that I think should heighten your concern about investing in Coinstar. Figure 8 is quite simple; it shows the association between quarterly reports of Coinstar's two principle sources of revenue - DVD rentals from Redbox kiosks and returns from the green coin changing machines - and quarterly Google searches for Redbox. I calculated the latter by averaging the quarterly search volumes in each state and then summing all states together (again using weights to account for differences in state population sizes).
As would be expected, there is no relationship between Redbox searches and coin machine revenues (red dots). However, there is a pretty striking quadratic relationship between DVD revenues and Redbox searches (blue dots and green line). That is, more people searching Google for Redbox is associated with more DVD rental revenue for Coinstar, but the marginal returns of each additional search diminish as the volume of searching grows large. I do not think this is good news for those thinking Coinstar will become a momentum growth stock.
There is another hugely important reason the relationship shown in figure 8 is important. It vastly enhances my confidence in the validity of the analyses in parts 1 and 2 of this article: it makes me believe that Google searches for Redbox are a viable indicator of how well Coinstar will do. It is the final nail in the coffin of my idea that Redbox would benefit from Netflix's lost customers.
Following the time labels on the graph, it appears that Redbox searches as well as DVD revenues increased over time (DVD revenues have come to dominate Coinstar's revenue totals). The increase in both appears to be slowing, which I think suggests that market saturation may be occurring. One reason I had invested in Coinstar was because I was excited about their new promotional campaigns. Figure 8 dampens my enthusiasm about that.
Figure 8. The association between revenues of CSTR broken out by DVD and Coin units and Google searches for Redbox.
Click to enlarge
Putting this all together, I have seen more than enough evidence to close out my long position on Coinstar. The revenues are highly dependent on the Redbox unit, which does not appear to be growing as robustly as it had in the past. They do not have massive sources of revenues from other units as yet (DVD revenues dominate the coin return revenues), and I see little likelihood that this will increase in the future. People are, after all, using less cash which should mean that they have fewer coins to roll. I think there may be interesting other projects underway at the company, but they are not my primary investment interest at present.
I think it is extremely risky to enter into a short position on CSTR given its high short interest ratio/days to cover. The fact that such a large number of the shares are short, I think means that we might expect a floor on the stock falling and a real powder keg that will explode on good news. This is too risky of a situation for me to short the stock. However, I do think the asymmetric bets available from puts on the stock may be worthwhile and I have bought a small number of April puts and I have closed my poorly researched long position. If it becomes apparent that Netflix customers are not opting for Redbox, I could see a moderate fall in the stock occurring.
As a final thought, I computed a regression of Coinstar's quarterly total revenues (from both the Coin and DVD units) as predicted by Google searches for Redbox and generated predicted values (I also included a variable that is the square of searches for Redbox to capture the quadratic relationship seen in figure 8). Figure 9 graphs the time trends in these predicted values and the actual total revenues. As can clearly be seen, this simple prediction has done remarkably well in the past. Worryingly for Coinstar longs, it also suggests a predicted slowdown in revenues in the coming quarter. I think we are not far enough into Q4 to make a reliable prediction, but what I have seen so far has made me reevaluate my position.
Figure 9. Trends in actual and regression predicted Coinstar revenues based on Redbox Google searches.
Click to enlarge
Additional disclosure: I am short both through puts.