Convolution Bounds on Quantile Aggregation

Jose Blanchet,Henry Lam,Yang Liu,Ruodu Wang

2024-09-06

Abstract:Quantile aggregation with dependence uncertainty has a long history in probability theory with wide applications in finance, risk management, statistics, and operations research. Using a recent result on inf-convolution of quantile-based risk measures, we establish new analytical bounds for quantile aggregation which we call convolution bounds. Convolution bounds both unify every analytical result available in quantile aggregation and enlighten our understanding of these methods. These bounds are the best available in general. Moreover, convolution bounds are easy to compute, and we show that they are sharp in many relevant cases. They also allow for interpretability on the extremal dependence structure. The results directly lead to bounds on the distribution of the sum of random variables with arbitrary dependence. We discuss relevant applications in risk management and economics.

Risk Management,Optimization and Control,Probability,Mathematical Finance

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the optimal value problem in quantile aggregation under an uncertain dependence structure. Specifically, given a set of random variables \(X_1, \ldots, X_n\) with known marginal distributions but unknown dependence structures, the paper focuses on the maximum and minimum values of the quantiles of the sum of these random variables \(S = X_1+\cdots+X_n\):

\[
\sup\{q_t(X_1 + \cdots + X_n) : X_i\sim\mu_i, i = 1, \ldots, n\}
\]

\[
\inf\{q_t(X_1 + \cdots + X_n) : X_i\sim\mu_i, i = 1, \ldots, n\}
\]

where \(q_t(X)\) represents the (left or right) quantile of the random variable \(X\) at the probability level \(t\in[0, 1]\). These problems have wide applications in fields such as finance, risk management, statistics, and operations research.

The main contribution of the paper lies in proposing convolution bounds based on Range - Value - at - Risk (RVaR). These bounds not only unify all existing analytical results but also provide new theoretical results. The convolution bounds are exact in many cases, easy to calculate, and have good interpretability, especially in terms of extreme dependence structures. Moreover, these bounds can be directly applied to the calculation of distribution bounds for the sum of random variables under any dependence structure.

The paper establishes new analytical bounds by introducing the convolution formula and proves that these bounds are optimal under certain conditions. These results are not only of great theoretical significance but also very useful in practical applications, especially in the fields of risk management and the economy.