I also asked this question about the graph below.

Why are the peaks asymmetric relative to the bottoms?

Why do peaks deviate a lot above the median but bottoms rarely deviate far below it?**

This is a deep question about Bitcoin’s bimodal distribution and the asymmetry of human behavior under gain vs. loss.

There are four primary reasons.

✔ Reason A — Positive feedback in bubbles is stronger than negative feedback in crashes

During bubbles:

new adopters flood in

leverage increases

liquidity increases

miners hold coin (reduced sell pressure)

global attention spikes

price reacts super-linearly

This causes overshoots above the power law.

During downturns:

long-term holders accumulate

miners must sell to survive (floor pressure)

increasing adoption continues

infrastructure continues to build

fiat liquidity re-enters gradually

These effects limit how far below the trend the price can fall.

Thus:

Upside momentum is unbounded.
Downside is bounded by real economic fundamentals.
✔ Reason B — Cost of Production sets a soft lower bound

Proof of Work creates a cost floor.
If price gets too low:

miners capitulate
supply decreases

difficulty drops

remaining miners mine cheaper

hashpower stabilizes

the network finds equilibrium

Because of this:

Bitcoin’s price cannot sustainably collapse far below the cost-of-production band.

This produces a lower asymmetry.

✔ Reason C — Human psychology is asymmetric

Economic behavior is not symmetric around gains and losses.

Mania produces exponential upside reflexivity.

Fear produces limited downside because strong buyers (HODLers) step in early.

In simple terms:

“Greed has no ceiling; fear has a floor.”

This psychological asymmetry maps directly onto price data.

✔ Reason D — The power law is multiplicative, not additive

This is the mathematical heart of the asymmetry.

In multiplicative systems:

Deviations above trend expand exponentially

Deviations below trend shrink logarithmically

This automatically creates:

Large, dramatic peaks
Shallow, compressed bottoms

This is exactly what your simulations show.