I was totally confused: Is randomness part of the process or can randomness in the process be designed?
Why did the question arise?
Earlier, I talked how my shift from reading online material (algorithm-driven) to in-person meetings (forced and random exposure) helped my investment performance in the last 6-9 months.
Then a couple of months ago, one of my investor friends asked me if I have looked at TVS Supply Chain. That triggered a chain of events which led me to Aegis Logistics which I found to be a very good business with an integrated value chain right from sourcing to distribution. Aegis turned out to be a great outcome.
The question of randomness became more pressing, the role of luck in investing became much more apparent.
If randomness improves outcomes, how do we design for it? I thought a lot about it and I kept going in circles – a process minimises random outcomes, randomness in process cannot be called a process. The confused-face emoji (with stars circling its head) came to mind.
Eventually, tired of going in circles, I asked AI. So, for the very first time here is the AI content for you:
(AI Content warning)
This isn’t really an either/or, and I think your own Aegis story already shows why.
Randomness and process aren’t competing for the same job — they operate at different stages.
Randomness is a sourcing mechanism. Process is a filtering mechanism. They’re not substitutes for each other; one feeds the other.
Look at what actually happened with Aegis: the friend’s message was pure randomness — nothing in your process would have generated that lead on its own. But once it arrived, your process didn’t get overridden. You still waited to check if last year’s blockbuster quarter was sustainable. You still listened to the concall before buying. You still capped the position at 2-3% until the moat was clear, then sized up only after @SHAIL_APTE insight resolved that gap. Randomness sourced the idea. Process still decided what to do with it.
The actual design move isn’t choosing between them. It’s widening the input funnel with randomness, while keeping the discipline of process on what gets through it. You don’t design a process for randomness — you’re right that’s structurally incoherent; a process needs repeatable rules and randomness by definition breaks repeatability.
What you design is your own exposure — meetings, conversations, people who message you unprompted — so more unscripted inputs reach a process that’s still doing its normal job of filtering once they arrive.
Put more sharply: randomness should never override process. It should only ever earn the right to be evaluated by it.
(AI Content ends)
What do you think of it folks?
Most investors spend their time optimising their process. What if the bigger edge comes from optimising what your process gets to see?
Can randomness complement process?
Disclaimer: This is not investment advice. It is a personal account of my own thought process and portfolio decisions, shared for learning and discussion. I hold positions in the stocks mentioned, which is a conflict of interest you should be aware of. Please do your own due diligence before making any investment decision.
