Looks interesting with revenue growth, stable margins, and high ROIC. This thing deserves a thread of its own, IMO. Maybe this is similar to MU, cyclical, long-term trend is upwards, buy when others are looking the other way?
SIMO’s memory controller business is a much better business than MU, imo, because there is less pricing fluctuation and it is fabless. Gross margins are fairly stable around 50%, so it’s not as good as a CPU or analog business, but much better than totally commoditized segments like memory.
I can start a thread and write a few lines, but my thesis isn’t deep: it’s just an OK to good and somewhat lumpy business that is cheep.
Every time I look at SIMO, I think "why is't this priced higher?". I first bought some shares of it in 2017, at $42. My "analysis" back then was (I wrote it down) : "Looks like a solid company. Makes money, has no debt, is in a market that has a future. Is the market leader?". Today I still agree with my thesis, and now it is priced at $39 ???
In a crazy market where "everything" is priced to perfection, I just don't get why this has been more or less flat for five years. My main reason for not buying some more SIMO is that often when I think Mr. Market is wrong, it turns out there is some valid reason for the price being what it is...
Not sure my comments will be that constructive but like you I look at SIMO periodically. Most times my thesis is about the same as yours. I've come to the conclusion that SIMO's just not growing very fast. Revenue has been near flat from 2017 till now. Margins have slimmed up a bit. I would bet they are looking for the next leap in growth. I hold MU as well. Sometime I wonder if SIMO is, in Pabrai's terms a "hidden compounder" If you purchased MU in the bottom of a cycle and held for more than 10 years MU compounds north of 20% per year through its cyclicality. I wonder if SIMO being a supplier of the semi conductor industry has similar traits over very long periods of time. My big thought would be what is the moat here? can it be sustained for 10 years? I dont know all the ins and outs of creating semiconductor controllers .