The theory equates the maximum output deviations (efficient frontier) caused by combined inputs with affinity-synergy in a system, which leads to a parametric volatility – a curve that is similar to data envelopment analysis. The input is a cumulative variable (e.g: merged assets) and the output is a flow variable (e.g.: merged incomes). Rather than being purely stochastic, volatility is estimated by a novel parameter for risk named synergy, which is constrained by critical input (scarce resources). The outputacceleration derived from the mergers among inputs, boosted by synergy, is the main foundation of the approach, which special case gives Shannon and Boltzmann-Gibbs entropies. Tests are done in the 11 USA Sectors over their quarterly financial statements, proving that synergy is significant for financial statements, whereas typical betas only present significance in stock market data. A practical application is a novel discount rate for valuation using synergy, whose results for each sector are stable and coherent with perceived risk. Systems that rely on causal relations between output and multiple inputs can be regressed under novel parameters, rather than reckoning exclusively in optimization procedures.