Not every "risk" is a risk
The two sided nature make it optimizable. Although, a quantitative optimization is difficult. It's a pity that not so much people know what risk is. Market risk is a risk but operational risk is a danger (wrong models, methods, calculations will usually not suggest a good decision). I've pointed this out here.
Five paradigms of market risk management
Duality - in your analysis, be aware that you have two parts: the known and the unexpected
Boundaries - try to find boundaries that distinguish situations where markets behave "normal" from the rest. Boundaries based on cost of risk should work.
Optimization - Risk optimization only works inside the boundaries.
Evolution - business is often co-evolution. Not surprisingly genetic programming is a general way to optimize
Game Theory - a mathematical study of uncertainty caused by actions of others
Risk is not uncertainty
You can bet on five possible outcomes of of a dice throw. It's still uncertain, but there's only low risk. A large portfolio of possible outcomes is very different from, a large risk.
Innovators, present something new…consequently they don't have much feed-back. Apple didn't know what people will say about the iPod…before they rolled it tout.
Innovators live in uncertainty for a while, but if they've built a technology stack their transformative innovation's placement will not be very risky...and real option valuation tools will help them to plan under uncertainty…
Market risk is two sided…but, as Daniel Kahneman, has demonstrated people show a tendency to strongly prefer to avoid losses to acquiring gains. A win doesn't feel as good, as a loss feels bad…five great assessments do not balance a bad one…?
This leads to risk aversion…and prevents market participants from finding the optimum.
There are many things that can be placed at a map of positive vs negative outcomes and consequently optimized (although. I name price as position at the value / cost (or external vs internal value) map…is there a socio-economic optimal pricing? This is even more difficult, when serving two-sided markets…actors and, say, system integrators…provide a process automation system to steel makers and steel plant engineers and contractors…
We've managed the problem for UnRisk by shipping solutions and development systems in one…under the same pricing and licensing scheme.
p.s. but there are objects of desire where you need to pick two of three things: "performance" (of systems) is three-sided if yo want high standards - high production - flexibility…you need to pick two.