Paper: arXiv 2309.01936
Authors: Hui Mi, Zuo Quan Xu, Dongfang Yang
Abstract
This paper investigates an optimal investment problem under the tail Value at Risk (tail VaR, also known as expected shortfall, conditional VaR, average VaR) and portfolio insurance constraints confronted by a defined-contribution pension member. The member’s aim is to maximize the expected utility from the terminal wealth exceeding the minimum guarantee by investing his wealth in a cash bond, an inflation-linked bond and a stock. Due to the presence of the tail VaR constraint, the problem cannot be tackled by standard control tools. We apply the Lagrange method along with quantile optimization techniques to solve the problem. Through delicate analysis, the optimal investment output in closed-form and optimal investment strategy are derived. A numerical analysis is also provided to show how the constraints impact the optimal investment output and strategy.
Complexity vs Empirical Score
- Math Complexity: 8.5/10
- Empirical Rigor: 3.0/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper employs advanced stochastic calculus, martingale methods, and quantile optimization to derive closed-form solutions for a complex optimal investment problem under tail VaR and portfolio insurance constraints. While it includes a numerical analysis, the absence of real data, backtests, or implementation details places it firmly in theoretical research.
Research Flowchart
flowchart TD
A["Research Goal:<br>Optimal DC Pension Management<br>under Tail VaR & Inflation Risk"] --> B["Key Methodology:<br>Lagrange Method &<br>Quantile Optimization"]
B --> C["Data/Inputs:<br>Stock, Cash Bond,<br>Inflation-Linked Bond"]
C --> D["Computational Process:<br>Transforming Tail VaR Constraint<br>into Solvable Optimization"]
D --> E{"Outcomes"}
E --> F["Closed-form<br>Optimal Investment Output"]
E --> G["Analytical<br>Optimal Investment Strategy"]
E --> H["Numerical Analysis<br>showing impact of constraints"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a closed‑form solution for the optimal investment problem of a DC pension member under joint tail VaR (TVaR) and portfolio insurance (PI) constraints, using a Lagrange method combined with quantile optimization techniques.
Source sentence
This paper investigates an optimal investment problem under the tail Value at Risk (tail VaR, also known as expected shortfall, conditional VaR, average VaR) and portfolio insurance constraints confronted by a defined-contribution pension member. We apply the Lagrange method along with quantile optimization techniques to solve the problem.
Assumes 6
- The inflation index follows the SDE dI(t)/I(t)=μ_I dt+σ_I dW1(t). market
Source sentence
To reflect the inflation, we define a stochastic inflation index process I, which is driven by the following stochastic differential equation (SDE): dI(t) I(t) = µI dt + σI dW1(t), where µI is the instantaneous return rate and σI > 0 is the volatility rate of the inflation index process.
- The inflation‑linked bond price follows dB(t)/B(t) = (r+μ_I) dt + σ_I dW1(t). market
Source sentence
The bond offers an instantaneous return rate r exceeding the inflation index return rate, so its price is driven by dB(t) B(t) = r dt + dI(t) I(t) = (r + µI) dt + σI dW1(t).
- All market parameters (μ_I, σ_I, r, etc.) are constant. model
Source sentence
We assume all the market parameters are constants.
- An admissible investment strategy π(t) is F‑progressively measurable on [0,T] with ∫_0^T ||π(t)||^2 dt < ∞ a.s. model
Source sentence
An investment strategy π(t), t ∈[0, T], is called admissible if 1One can also assume that T ′ follows an independent distribution. It is an F-progressively measurable process on [0, T] such that Z T 0 ||π(t)||2 dt < ∞ a.s.; 2.
Show 2 more
- The death time T′ of the member is a known constant larger than the retirement time T. other
Source sentence
Assume the death time of the member T ′ is a known constant bigger than T 1.
- Assumption 3.1 (well‑posedness of the benchmark problem) holds, i.e., there exists λ>0 such that certain integrability conditions are satisfied. statistical
Source sentence
The following requirement is minimum and henceforth assumed. Assumption 3.1. Assumption 3.1 holds if and only if there exists some λ > 0 such that E[ρ(U ′)−1(λρ)] < ∞, EU�(U ′)−1(λρ) < ∞. (3.4) Proof.
Data
No dataset named in the text.
Costs
Not applicable
Finds 4
- The optimal terminal surplus Z*(T) is piecewise defined across four market regimes, combining shifted Merton‑type strategies and a minimum guaranteed level from the PI constraint. theoretical
Source sentence
When R(z) < z < C(z), the optimal terminal surplus Z∗(T) for the problem (3.2) is Z∗(T) = I(νρ) ρ ⩽ρ, I(νρ) ρ < ρ ⩽ρ, I(νρ −λν α ) ρ < ρ ⩽ρℓ, ℓ ρ > ρℓ, and the corresponding optimal terminal wealth for the problem (2.5) is X∗(T) = I(νρ) + L(T) ρ ⩽ρ, I(νρ) + L(T) ρ < ρ ⩽ρ, I(νρ −λν α ) + L(T) ρ < ρ ⩽ρℓ, ℓ+ L(T) ρ > ρℓ, where I(x) = (U ′)−1(x), ρℓ= U ′(l) ν + λ α, ρ = ρ + λ
- Numerical results show that adding the PI constraint to the TVaR problem lowers the lower bound ℓ and significantly reduces tail risk in adverse market scenarios. performance
Source sentence
We can see the additional PI constraint can improve the distribution of Z∗(T) by setting a lower bound ℓ, which implies it significantly reduces the tail risk in the worse market scenarios.
- A higher confidence level α (less stringent TVaR) leads the member to allocate a smaller proportion to the stock and a larger proportion to the inflation‑linked bond. performance
Source sentence
A higher α means less effort that needs to pay by the member to achieve his goal at retirement. In the meanwhile, there is an uptrend in the proportion of the inflation-linked bond.
- Increasing the TVaR reference level z makes it harder to satisfy the left‑TVaR constraint, prompting higher stock allocation and lower bond allocation. performance
Source sentence
When z increases, it is more difficult to satisfy the left TVaR constraint. As is shown in Figure 7, the stock as a percentage of total wealth has becoming larger as z increases, and the bond is just in the opposite trend.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 13 related papers). Coverage is partial, so this is neither replication nor refutation.
1 paper in the archive
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.