*** Way too complex = way too fragile *** * Complexity of par | ChainHedge

*** Way too complex = way too fragile ***

* Complexity of particular farming\trading strategy (advanced combination with mix of platforms etc.) decrease system robustness significantly.

Usually we can't even guess what may go wrong. Even if we can assume vectors of attack, we can't count probabilities accurately. When I see some kind of 3-4 steps strategies, I assume it is the fastest way to fuck up in the long run. Most of people have no competences to even measure the risk for particular strategy. During the time when "too much money in the market" these strategies will lead to *risk adjusted* negative APY, even it shows as double digits juicy one. You have to study really well how things work and even go deeper in code if it is about automatized strategy which is not trivial task for anyone.

* Same way additional optimization of funds allocation will lead to decreased robustness of whole system.

The "smartest" people in the room are not who made the most return for bull market, but who survived after. All the over-leveraged\highly optimized yield fools are out of the game eventually.

* The most reliable "stress test" for immutable smart-contracts is time.

Then longer something works than higher chances it will keep working. Decent audit(s) is a must, but everyone can make mistakes, it is just human factor. The real money on the table make it stress tested much better than any audit.

For now:
Complexity of the most systems increased significantly. Even Maker began its degen game with Real World Assets(RWA) which *may* eventually lead to total disaster and failing of DAI system as well as every liquidity pool which includes it. For me it looks worse than USDT since Maker will not be able to handle compliance requirements. So even Curve 3pool is risky for me right now without providing any reasonable compensation for such risk.

I don't assume we have to avoid uncertainty and avoid taking high risk, but at least it should be properly compensated. When during last bull run we went to farm something new, we did it at least for triple digits APY with relatively measured time limited risk.