Mohnish Pabrai: Why I Sold Alibaba for Tencent - Capital Allocation Engines

Pabrai's investing signal framework: 'There's very strong signal value when someone buys something there is very weak signal value when someone sells something so in general if you're looking at Buffett's portfolio and we see that he bought something that is very important to pay attention to but when he sells something that is not as relevant because we buy things only for one reason okay we buy things to make money...we sell things for 100 reasons.' This explains why sells shouldn't be over-analyzed - could be tax loss harvesting, rebalancing, portfolio constraints, etc., not necessarily negative business view.

Alibaba sale rationale: 'One of the reasons I sold Alibaba was for tax loss harvesting...I had studied and learned about Tencent and Prosus after we had made the Alibaba investment and after we made the Alibaba investment the stock went down and Prosus and Tencent also went down but we hadn't invested in it yet.' Execution: 'If I sold Alibaba we got the tax loss harvest' while avoiding wash sale problem where 'you can buy back the same security after 30 days and in the 30 days it can move...against you.' Solution: 'I sold Alibaba I moved the proceeds to Prosus...not Tencent because Prosus is even better than Tencent and I never need to buy Alibaba back again because I thought the Tencent business was superior to the Alibaba business.' Result: 'Killing multiple birds one stone where I'm moving from a business that's good to a business that's even better I get the tax loss harvest and I would have been willing to buy Prosus at the higher price.'

Why Tencent is opaque: 'It's a mystery because they don't really explain the business and the reason they don't explain the business is because they don't really want attention they don't want competitors to understand the mousetrap they have they definitely don't want governments to understand it they really don't want anyone to understand it they just want to be in stealth mode.' Amazon AWS analogy: 'Amazon was like that with AWS and when the whale finally surfaced they instantly attracted two large competitors who are going at it head to head with them.' Strategic rationale: 'I can understand why they don't want to talk about it...they feel they will just attract other people into the field...they just don't see any upside and a lot of companies do that a lot of companies have a kind of balancing act between what the SEC requires and what disclosures they want to make because all the competitors are reading all their annual reports.' Pabrai's research: 'I was able to figure out their business and what it's all about really because of comments that Koos Bekker at Naspers made.'

Tencent's dual business model: 'They have an army of software engineers and this army of software engineers is incredible...they can produce a wide range of things which are very valuable in this digital economy...they've chosen in the past to direct that army towards video games messaging...now with AI and with cloud.' First engine returns: 'When Tencent puts money into its digital army historically they've generated 65 percent annualized returns on that investment which is easy to understand...if you put out a video game and suddenly a billion people are using it you're going to make a killing.' Capital deployment problem: 'It cannot use the capital it generates so it's like the MasterCard or Visa or Amex problem...or like Google search engine generates a lot of cash cannot use it.' Scale limits: 'Every billion dollars they spend allows them to hire about 4,000 software engineers...once you get to hiring 10, 20, 25,000 software engineers net in a year...you start running out of places to hire them from.' Second engine: 'When they cannot use that money anymore then they take the rest of it and send it to their second engine which is the second bazooka which is investing in early stage businesses and private deals...there they have a really good track record they've historically done about 35 percent a year which is exceptional.' Focus: 'Tencent only wants to put money into software guys nothing else.'

Capital allocation comparison: 'There are only two companies out of all the large tech giants who are really good at capital allocation and one is Amazon and the other is Tencent and Amazon is not as good as Tencent because they put a lot of money into non-digital businesses so Tencent only wants to put money into software guys nothing else.' On Alibaba: 'Alibaba is not as digital as Tencent Alibaba's capital allocation engine is more sporadic it's more ad hoc it's like kind of like Microsoft or Google where they have a lot of cash once in a while some deal comes up they do it Tencent doesn't work like that Tencent deliberately takes all the cash they generate and pumps it out.' Competitive advantage: 'If you have a rapid redeployment engine which Microsoft doesn't have...Google doesn't have and most tech businesses don't have it gives you a huge edge so for me it was I preferred that bet and so I made the switch.'