 One of the most requested stocks for me to analyze was, of course, Micron, especially as it went from above 60 to what the current vertige something, and except for the stock decline, there is something else that's very attractive to many of you, and those are the fundamentals. We are talking price earnings ratio of 2.5, forward price earnings ratio of 3.5, and the price to tangible book value is at one. So every value investor thinks this is a screaming buy. However, whenever there is something of value, it might also be a value trap, and when you deal with such stocks, it's better to go to the persons that know very well the company that have known the company for more than 15 years, than know how the company breeds, and a person like that is Yao Kai. You might know him from the InvestWit JYK YouTube channel. He is a computer scientist geek. Let's call him graduate from Cambridge, computer science, software, data processing, everything you want to know from the sector, better go to a specialist, and he is definitely a specialist on Micron. And to make the story even sweeter, he has been short the stock for the past six months. He has even been telling me to short it for the past six months. I fortunately didn't go into shorts yet, so I didn't listen, but he has made the killing for now, and I'll just talk to him about the company. What are the risk rewards? What can an investor expect in the long term, in the short term? Should it be short? Should you sell? Should you buy? In this video, we will have a perfect idea of what the company is, what are the risks, and what are the potential rewards. So welcome Yao Kai. Thanks you for doing this. This is our first video collaboration. I think I hope there will be many, many more. Thank you. Glad to be on. So let's start immediately with the company. Let's first tell us, I think everybody wants to know, why have you been short micron, which is a company that has a price earnings ratio of 2.5 and a price to book ratio of just one. Everything is tangible and practically no debt. So why were you short? So I started shorting when the stock price was 42, and then I kept on adding my shorts until 60, just before it started turning. The reason was back then when I started shorting, it was at 2.5, I think, PB-ish. And micron is one of those cyclical companies that you really can't use PE to value. And they were the simplest, in a nutshell, I shorted it because their days were too good. They were making too much money. And there is such a thing as making too much money for cyclical companies because the capital always will flow to an area where you have excess return. And excess return, in this case, is memories. And they were getting 50% return on equity and 40-something percent return invested capital. And those are non-sustainable numbers unless you have some kind of moat. And in this case, they don't really have a moat because their moat is their know-how. They know how to use these machines, but these machines were not built by them. They were built by lamb research, applied materials, people like that. So when memory was selling at such high margins, everybody will go in and increase capital expenditure. And two years down the road, the price would decline, and that's what happened. And obviously, there is also the side of demand because as memory gets cheaper, people like us would use more memory to get better performance. And for example, you could have the move from, say, Hadoop to Spark, which is a memory. They're approximately the equivalent. One is a memory. One is on disk. On disk is much cheaper. But if memory is much cheaper, suddenly much cheaper, then you can use more of it to get more performance. And when memory is going the other way, it's getting very expensive, then we start coming up with hopefully clever ways to reduce memory consumption. So there is an effect on the demand as well. So my prediction was that the earnings will collapse. And together with that, the price would collapse as well. And then the earnings would—I've heard people say, oh, if they cut their earnings by half, they're still at, I don't know, something like a 65 PE or something, right? If we look at the forward price earnings from Morningstar, it says 3.5. Right. If you cut it by half, it will be only 7. It will still be a good value. But no, this thing can lose money. It's not that they will lose 50% of their earnings. It's very likely that we'll lose more than 100% of their earnings over a year or two. So let's say in 2021, they will probably have negative earnings. Earlier than that. I think this year or next year, 2020, probably, they will have negative earnings. And they—correct me if I'm wrong—they already lowered their guidance in 30 years. Yes, they lowered their guidance quite a bit. Okay. Okay. So these numbers are just for now, and they will turn worse. And we don't know how much worse. Right. And most of the time, they turn negative. Like on every downturn, I think four out of five, they go negative. Okay. And then you look at something that went from like a four PE to negative something PE, right? So. Okay. And now that the stock is from 60 to 40, are you still short? I'm about to close my shorts. That was my price target. And the reason around 30 was my price target was it was going to get close to book by 30. So I bought these puts for 2020 expiry. And they were making something like 50% ROE. So within a year and a half or so, they would get something like 35 or more book value even considering the lowering of their ROE. So the reason I think one PV is a good anchor is that this business has inherent value. Are we going to need memory in the future? Yes. Memory in the future. And does it cost any less to make these, these equipments in order to produce that memory? It doesn't cost any less. In fact, any, every time we get a technology upgrade, it costs more and more, right? So the replacement value is a good proxy for the long-term value of this company. And when I started shorting was 2.5 PV and when it was highest was 3 PV and now it's one. And I'm considering to close my shorts. But for example, then if we look at, if we look at from a value investing perspective from the fundamentals, it then should be, let's say you buy, because I'm practically buying this equipment for one somewhere in down the line that should make money or not. You should, you should, if you buy it now at one PV and hold it for long term, you should make money, but you will make average money because that's a replacement value and the company's long-term value is their replacement value. But as a value investor, I don't want to buy at fair value. I want to buy at discounts. So if this thing goes to 0.6 PV or something, I will buy it. But at the moment, since they're at their long-term fair value, I think I'd rather sell because the easy money is going from two plus PV down to one PV, not from one PV down to 0.6 PV, because who knows what's going to happen then? Because Samsung already, Samsung and Heineck's both came up and said they will curb capital expenditure because the price has been collapsing. And that would have an effect a year and a half to two years down the line. So there might be another bump in memory prices then, but it could also coincide with the recessions and whatnot. Ignoring that, you already see the reaction from an industry participant, except the Chinese, which are not here for the money, they're just here because they want memory to be produced in China. So you have that wrench, but the easy money, I think the easy money on the short side has been made. Maybe at some point you will get a chance to make the easy money on the long side. All right. So what are your expectations for MicroN as a company? We are now talking about the value investing. Is there any chance for, let's say, long term that it becomes a value trap? Can it go bust? It can go bust, but at the moment it's not likely. It's a lot less likely to go bust compared to, say, 2002 and stuff like that. They were very close to bankruptcy back then, but back then the industry was spending 58% of sales in capital expenditure. So that is crazy, crazy stuff. When you have that kind of expenditure, yeah, you rack up a lot of debt because that's the only way you can afford it. And you might just go bust, but at the moment they're spending some like 30%, 38%, which is a close, slightly above long term average, but not that much higher. So I think their liabilities are just three billion long term. Yeah, liabilities now is very low, but for these companies, usually it's when they do a huge capex, they would burn up a bunch of money, take up on a lot of debt, and then that's when they get in trouble. And when is the next capex cycle for the new technology? Currently we are at, for RAM, we're at 1Y is basically the node size. And the next one's probably going to be seven something, so seven nanometers. But I think that will come within three years, two to three years. Because usually that's the Moore's law, remember that it's like every 24, what Moore's law says, 24 hours you double the transistor counter. Then the way you achieve that is by going to a new technology node. And the next one probably within that. MicroN just released their 10Y, which is the second generation of the 10 nanometer node, and that's this November. So maybe within two years everybody will have to spend for the seven nanometers. And these machines that they bought from LAM research and things like that, they cannot be used for the new technology? Most of them, no. Some of them maybe, because for manufacturing you need the wafers. You need the things to etch it. So you use laser beams, you cut things on the wafer. And then you have to make the mask yourself in order to do the cutting. So the etching, sometimes you can reuse across generations. The masking, the mask you have to make every time different. And there are a bunch of other things that are pretty crazy that they have to do that also can't be used. So a lot of them will lose their value, but not everything. And then the actual manufacturing factory, the factory itself, because you need a dustless environment and all that stuff. That stuff, obviously you can reuse. That's not cheap. It's very expensive to build just the thing to contain the machines that you use. So that explains, when I look at their balance sheet, the gross asset value is 50 billion, and then it's depreciated. And the net value is actually just 24 billion. Yeah, yeah, the depreciation is going to be very high, except the room. The clean room is, yeah. But still the depreciation is possibly going to be very high the next years. And the capex is also going to be high. So we shouldn't expect any dividends in the next year or two. Yeah, they've never, I don't even know if they've ever paid dividends this company. No, in the last 10 years they didn't pay anything. Yeah, they've never been able to pay any dividend. It's not one of those steady cash flow consumer defensive notes, it's it. And they also increased their number of shares by 50% over the last 10 years. Yeah, because they were going to die. They were so close to bankruptcy that they were getting bought by a Chinese company, except they were stopped from doing that because Americans don't like Chinese having technology around memory. So yeah, they were in a very tough spot. The whole memory industry were in very tough spots like repeatedly. If you look back to 80s, Intel got messed up because the Japanese went into memory. Intel used to be a memory producer, and the Japanese went in there. And Intel got kicked out of the market because they couldn't make memory cheap enough. And then the Japanese all died in around 2001 because that's when the Koreans got in the game. And the Taiwanese also got in the game. So basically we can say the moat is there is no moat in this business. There's very little moat in this business. Okay, okay, okay. So as I think many of our viewers on our channels are either mostly long micron because wherever I read, micron is cheap, cheap, cheap, cheap, undervalued, book value, screaming value. So most of our viewers would be, I think, long now. And unfortunately long down from 60 probably, because at 60 it was still a price in its ratio. Yeah, so something like that. So what would be the strategy with a company like this, long or short for somebody that's already long, or should they wait it out, or how do they manage the risk reward, let's say? I wouldn't consider too much my entry price because that's happened, that's already happened. I mean, the choice is do you want to have micron in your portfolio or not have micron in your portfolio? For one PB, I wouldn't have micron in my portfolio. I wouldn't short it either. I just don't touch it because it's going to be average return. And if I want average return, I could just get S&P instead of micron, much lower volatility and less to think about. But when they are above two PB, that's usually something is dangerous. Because everybody else will go into the high return on capital? Yeah, because at that time their return on capital will be insane, it will be 30%, 40%. Things that Coke can't even do for example. On the down cycle, PE will be non-existent, usually they go negative. PB would be 0.5, 0.7 or on a mild down cycle, they reach one PB and then they kind of bounce off one PB. So I would say if you can find this thing at 0.5, buy some as a pretty good risk reward, they can go bust. So you can lose 100%. But when they do good, they give you 10X within years. That's the most important thing to understand. The risk is there, but when the cycle turns, then they could really do it wrong. Yeah, when the cycle turns, these things are explosive. Because the main mode really is time. Because even if you want to build a factory, it's going to take you two years, one and a half years at least, build a factory, put the machines in. You have to tune these machines. They don't arrive out of the box, you click a button and it works. You have to tune these things, you have to set them up, you have to train the people to operate them. So it would take time. Time is their biggest mode. But your mode only lasts for a year and a half and two years. After that, new things will come online. Yeah. So it's very interesting because Seth Klarman I think was the largest owner of Micro in 2016. Yeah, it did really well in 2016. In fact, I missed that run, I was still kicking myself. I should have bought, but instead of buying Micron, I bought a bunch of SSDs. I bought tens of thousands of dollars of SSDs and I made money on the SSDs, yes. But I could have made 5X on Micron. And you were spending your time driving around your Porsche, I think that you crashed later, but okay. I no longer have a fancy car. I live a plebeian life now. Okay, okay. So that's why you missed on Micron the way up. But I think Klarman was buying at 10 and he sold around 40. So he did pretty well in 2016. Yeah, I think that's the way to play it. You get it close to book or under book. And if you're able to spot the CapEx trend, so you look at CapEx and overlay that with revenue, if the CapEx as a percentage of revenue is very low for consecutively two years, then you have a shortage. And then when you have a shortage, the price will shoot up. And then the earning will shoot up. But if their CapEx is too high, total, not just them, the entire industry, good thing is you only have three big players now. Before you had like six or seven, you have to look at. Now there's only three. So you look at three, find their annual report, say, oh, how much money you spent on fabs, okay, lots of money, lots of money, everybody spent a lot of money, but like Samsung in 2017 alone, I think 17 or 18, spent as much money in CapEx as Micron's market cap. So they basically built another Micron. Yeah. Right, so like. And how does the recession impact all these companies? I mean, I think that's just a little risk to keep in mind. Yeah, recession will make the model go to really bad because a couple of things. Well, I would like to discuss why they are like actually cyclical because their Intel is not a cyclical of these guys. They're all semiconductors. But the reason they're cyclical is do you know if your phone's memory chip is made by Micron? No, you have no idea. Do you know that when you get a laptop, the memory chip is made by Micron, Samsung or some Chinese factory, you have no idea because you don't care. There's no branding in here. There is no pricing power whatsoever. You sell at the lowest price everybody else is willing to sell at, right? But it's different from Intel. You say, oh, I have an i7 CPU. Maybe it doesn't mean anything to you, but you've heard of Intel i7. Intel inside should be good. You don't have an Intel PC. You don't know I have a Micron PC. So that's the difference. So they will be super cyclical. And in a recession, people stop changing their phones, which I think is close to 20, 30% of the DRAM and NAND market. So people use a phone for two and a half or three years, instead of a year and a half or whatever. Then your demand gets cut in half. And people don't replace their PCs, even though their PCs are less and less. And the other side is data centers. People don't spend much on data centers either. Everybody is tied on money. Everybody cuts CapEx. And the story goes, yeah. So they will be really screwed in that. So what should we conclude? So you should really be careful and manage your portfolio exposure if you want to be exposed to this. On the short side, perhaps, as you said, the easy money has been made. On the long side, if you can watch a recession, if you can watch the CapEx spending and when the memory prices might shoot up in the next two years, then you might catch the run-up. Yes, yes. You have to predict the memory price. Yeah, yeah, yeah. Which is something that probably 99% of people out there cannot do. Right, which is why I puzzled why people are so into this kind of stock. Semiconductors is, I mean, I was- Because it was at the price earnings ratio of seven. Yeah, you can't look at that. It's too simplistic. What do you think? Price to sales, EB to sales, price to book, these are all better indicators. Yeah, great, great. I think I hope we have helped a lot of people with this conversation. Each one of my viewers can drop to your channel because there are a lot of great, I must say, gem videos on your channel. So a lot of technical details. So everybody loves that can drop to your channel to see that. I'll put the link also in my description for your channel. I'm sure Yaw Kai will put it for my channel. So your viewers might also want to drop off to my channel. My 900 viewers. Doesn't matter. As long as I think you are so technical and so detailed that of the 900, one will be Klarman, one will be Munger, one will be Buffett, you have me. And then we'll see who are the others, 896 we must see. Well, thank you. And you were one of the first subscribers to my channel. And yeah, and I really- And I think I was one of the early subscribers to your channel. I just remember you- The early subscribers and I think you were, I think the first five subscribers to my research platform. So that was also- I think I really enjoy a lot of the information you provide, especially on mining, because that's, I think you were the person that initiated me into resource investing in terms of like the mining industry. And I think it's close to- I'm gonna move your nerves on, thank you very much. You're welcome, you're welcome. So, but it's good that you got that tip and paid just one coffee when you came to the Netherlands, but that's a different story. Let's talk after the video about those things. So thank everybody for watching. Thank you. Check the links and we'll keep you updated on Micron and many other companies. Click like if you like this interactive video content. I'm sure you will do that and we'll make many more, right, Jaukai? Sure. Perfect. All right. Thank you and see everybody in the next video. Thanks for having me.