The shift toward modelability and measurability really stands out here. A lot of value folks still lean on qualitiative narratives rather than KPI tracking, but framing thesises with explicit time windows and quantifiable drivers (like Tetlock's forecasting principles) seems smarter long-term. One observation though: your concentrated book (70% in top 5) paired with 7.5x HHI implies heavy overlap or sizing risks. When positions like CHTR hit volatility despite solid fundamentals, how do you balance conviction against the psychological strain of carrying outsized exposure during extended slogs?
Hey Neural, that's right, the book is deeply concentrated. As someone who hangs their hat on deep fundamental stock selection, I don't see much point in running a book of 50 or even 30 stocks, nor could I maintain quality research at that scale as an individual. I've found 10-15 individual stocks to be about right, generally at the upper end of that range, which tends to translate to a ~8-10x HHI.
Regarding psychological fortitude, I just haven't really felt psychological strain from position drawdowns in the past. I think the cure to strain is having already done all the fundamental research. Clearly, no one's actually happy when some prominent position in their portfolio is getting clobbered. But, given that the fundamentals are good, the drawdown isn't a bad thing as long as you don't have LPs or lenders pounding the door for money.
The bigger question in my mind is when and how do positions become sized up to these very high levels? The answer (to me) lies mostly in considering the distribution of outcomes, both in its shape and in its definability. Every now and then, we can find a situation that offers a great way to track our thesis and where we are able to really understand the skew/kurtosis in the underlying distribution that our price target represents.
We get to the end of our analysis and think, "Okay, we can definitely lose money if XYZ happens, and that probably means we lose 10 or 15%, but we'll know when and why, and we can sell. And then in good outcomes, we make 40 or 50%, so that's a great r/r." If we can confidently get to that sentence and see that we definitively understand the risk pool, that the cost of being wrong is palatable, and that the thesis is measurable, that's going to be a very large position for me. Those are sort of rare ducks, though. We felt that way about CYND before. Currently, we're holding a 20%+ position in an event-driven situation that we feel that way about.
On the opposite end of the spectrum, something like China Yuhua Education that we wrote up earlier in the year was one where we got to really know the risks and catalysts, and the upside was clearly tremendous, but the downside was really high, and frankly, roughing out tail risk probabilities was just really hard. That felt like a great r/r, maybe even better than the top-weighted positions, but given the magnitude and likelihood of significant losses, it almost has to be one of our smallest positions, no matter the upside.
Most positions fall in between these extremes. Charter was like that. Neogen was like that. Dowlais was like that. And then sometimes those names get walloped one day for a very understandable reason (like NEOG's CFO getting fired), and you get a chance to scale into your risk because the skew suddenly gets a lot better, and it was only moderately sized before.
So, this is a long way of saying it's complicated. We want to bet big on things where we have the highest degree of thesis tracking and that we simultaneously think are unlikely to take a complete drubbing. Finally (and obviously), we also need to be extremely mindful of our broader factor exposures, because if two top positions end up highly correlated with some random factor (China regional risk) and have only a minimal stub of idiosyncratic risk, that's just a mess.
The shift toward modelability and measurability really stands out here. A lot of value folks still lean on qualitiative narratives rather than KPI tracking, but framing thesises with explicit time windows and quantifiable drivers (like Tetlock's forecasting principles) seems smarter long-term. One observation though: your concentrated book (70% in top 5) paired with 7.5x HHI implies heavy overlap or sizing risks. When positions like CHTR hit volatility despite solid fundamentals, how do you balance conviction against the psychological strain of carrying outsized exposure during extended slogs?
Hey Neural, that's right, the book is deeply concentrated. As someone who hangs their hat on deep fundamental stock selection, I don't see much point in running a book of 50 or even 30 stocks, nor could I maintain quality research at that scale as an individual. I've found 10-15 individual stocks to be about right, generally at the upper end of that range, which tends to translate to a ~8-10x HHI.
Regarding psychological fortitude, I just haven't really felt psychological strain from position drawdowns in the past. I think the cure to strain is having already done all the fundamental research. Clearly, no one's actually happy when some prominent position in their portfolio is getting clobbered. But, given that the fundamentals are good, the drawdown isn't a bad thing as long as you don't have LPs or lenders pounding the door for money.
The bigger question in my mind is when and how do positions become sized up to these very high levels? The answer (to me) lies mostly in considering the distribution of outcomes, both in its shape and in its definability. Every now and then, we can find a situation that offers a great way to track our thesis and where we are able to really understand the skew/kurtosis in the underlying distribution that our price target represents.
We get to the end of our analysis and think, "Okay, we can definitely lose money if XYZ happens, and that probably means we lose 10 or 15%, but we'll know when and why, and we can sell. And then in good outcomes, we make 40 or 50%, so that's a great r/r." If we can confidently get to that sentence and see that we definitively understand the risk pool, that the cost of being wrong is palatable, and that the thesis is measurable, that's going to be a very large position for me. Those are sort of rare ducks, though. We felt that way about CYND before. Currently, we're holding a 20%+ position in an event-driven situation that we feel that way about.
On the opposite end of the spectrum, something like China Yuhua Education that we wrote up earlier in the year was one where we got to really know the risks and catalysts, and the upside was clearly tremendous, but the downside was really high, and frankly, roughing out tail risk probabilities was just really hard. That felt like a great r/r, maybe even better than the top-weighted positions, but given the magnitude and likelihood of significant losses, it almost has to be one of our smallest positions, no matter the upside.
Most positions fall in between these extremes. Charter was like that. Neogen was like that. Dowlais was like that. And then sometimes those names get walloped one day for a very understandable reason (like NEOG's CFO getting fired), and you get a chance to scale into your risk because the skew suddenly gets a lot better, and it was only moderately sized before.
So, this is a long way of saying it's complicated. We want to bet big on things where we have the highest degree of thesis tracking and that we simultaneously think are unlikely to take a complete drubbing. Finally (and obviously), we also need to be extremely mindful of our broader factor exposures, because if two top positions end up highly correlated with some random factor (China regional risk) and have only a minimal stub of idiosyncratic risk, that's just a mess.
Really love this answer.