Feedback Loops

“The first rule of compounding: never interrupt it unnecessarily.” Charlie Munger’s line about compounding is really a line about feedback loops — a system whose output loops back around to become its own input, so each cycle builds on the last instead of starting from scratch. Reinvested earnings compound into more earnings; interrupted compounding just resets the clock.

Positive and Negative Feedback Loops

Positive feedback loops amplify change, pushing a system further from where it started — sometimes toward growth, sometimes toward collapse. Rising stock prices generate enthusiasm, which drives more buying, which pushes prices higher still, until the cycle either finds new equilibrium or becomes a bubble.

Negative feedback loops do the opposite: they pull a system back toward stability. A thermostat corrects toward the setpoint; a body sweats or shivers to hold its temperature steady. Markets correct overvalued stocks the same way — the loop exists specifically to undo the deviation, not amplify it.

Feedback in Learning

Elon Musk has called a tight feedback loop — constantly reviewing what you’ve done and asking how to do it better — the single best piece of advice he can give. It’s the same principle applied to skill: deliberate practice works because each attempt’s outcome becomes the input for the next attempt. Remove the feedback — practice without ever checking the result — and the loop breaks, and improvement stalls with it.

Feedback in Business and Investing

Berkshire Hathaway’s own structure is a positive feedback loop by design: earnings from acquired businesses get reinvested into acquiring more businesses, which generate more earnings, which fund the next acquisition. Munger and Buffett didn’t stumble into this — they built a system specifically to keep that loop compounding for decades without interruption.

The negative version shows up just as reliably: overvalued stocks eventually face downward pressure that returns prices to more sustainable levels, the market’s own correcting mechanism doing exactly what a thermostat does.

The Time-Lag Problem

The hardest part of working with feedback loops is that the effect often arrives long after the cause, which makes the connection between them easy to miss. A bad habit’s health consequences, or a policy’s economic effects, may not surface for years — long enough that the actual cause gets forgotten and the wrong lesson gets learned instead.

The Broader Lesson

Most complex systems — economies, markets, habits — are governed by loops like these, often several running at once and interacting in ways that are hard to predict in advance. The useful skill isn’t predicting the outcome; it’s recognizing, while still inside the loop, whether it’s the amplifying kind or the correcting kind — and, per Munger, not interrupting the good ones by accident.

Sources

• ”The first rule of compounding: never interrupt it unnecessarily” — attributed to Charlie Munger; widely quoted
• Feedback loop as “the single best piece of advice” — Elon Musk, interview remarks (widely circulated; no primary transcript located), paraphrased here rather than quoted