YQYunzhe Qiu

Yunzhe Qiu

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Yunzhe Qiu has a total of 33 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Dynamic Inventory and Derivative Control Under Tail Risk: An Application to Hog Production

    by Panos Kouvelis, Ye Liu, Yunzhe Qiu, Danko Turcic

    Undiversified agricultural processors risk severe losses when unhedgeable biological yield drops coincide with commodity price swings. We study a hog processor that jointly manages live-hog inventory and exchange-traded financial derivative contracts under Dynamic Conditional Valueat-Risk (DCVaR). Each week, random cohorts of underweight and regular-weight hogs finish growing; the processor routes them among a fixed-quota meatpacker contract penalizing underdelivery, cash spot markets, and finishing barns, while trading futures and option contracts on hog, pork cutout, and feed prices. Existing planning tools fit this problem poorly: risk-neutral models leave the firm exposed to price-and-yield crashes, whereas rolling intrinsic policies replace random arrivals with point forecasts and under-buffer contract quotas. Because tracking two cohorts, autocorrelated prices, and multi-contract portfolios makes exact dynamic programming intractable, we prove structural properties of the optimal policy: tail-risk aversion links basestock targets to the financial hedge and can induce mature-hog retention even when a risk-neutral firm sells out. Building on these properties, we propose a sharp, closed-form heuristic that pairs a quadratic program for futures and options with risk-adjusted newsvendor thresholds; this policy characterizes when contracts substitute for or complement barn inventory and outperforms existing tools. On proprietary farm and exchange data, our heuristic comes within 5.0% of an empirical dual upper bound, matching deep reinforcement learning (4.4% gap) within 0.6 percentage points and beating rolling intrinsic optimization (15.5%); against a sell-all baseline (21.3%), routing closes 12.4 percentage points and financial hedging 3.

  • Open Access

    Government Loan Guarantees for the Newsvendor: Credit Incrementality, Upstream Incidence, Default Risk, and Credit Signals in Supply Chains

    by Seung Hwan Jung, Panos Kouvelis, Yunzhe Qiu

    Government loan guarantees support SMEs worldwide, but program evaluations usually treat them as bank-firm credit products and ignore the supply chain context. A strategic supplier can adjust the wholesale price in response to a guarantee, and demand uncertainty adds operational risk absent from standard guarantee models. We incorporate a government guarantee in a newsvendor supply chain with a Stackelberg supplier, a capital-constrained retailer, and a competitive bank. Where pass-through is positive, a stronger guarantee raises the wholesale price, and the supplier captures part of the relief, while the retailer can be worse off than under no guarantee. Default risk can rise when lower borrowing costs induce higher borrowing amounts. At high coverage, loan spreads can approach zero even as default risk remains positive, making spreads a poor proxy for borrower or government risk. Furthermore, the government guarantee alone need not materially raise supply-chain efficiency. In the numerical analysis, higher demand variability and coverage increase default risk and expected government payments. Retailers with low working capital generate more incremental credit and larger expected government payments. Partial trade credit can raise efficiency and reduce government exposure simultaneously. A uniform coverage ratio has heterogeneous fiscal consequences. These results suggest that differentiated coverage by borrower wealth and demand risk is desirable.

  • Open Access

    Welfare vs. Revenue: Comparing Government Loan Contract Designs for Green Production

    by Jiyoung Han, Jinwoo Kwon, Seung Hwan Jung, Yunzhe Qiu

    Government loan programs for green production differ from standard subsidies in that public funds must be repaid, creating a fundamental tension between promoting output and preserving fiscal resources. We study how a government should design loan contracts for a financially constrained firm by comparing two policy objectives: welfare maximization and revenue maximization subject to a minimum green output requirement. These objectives yield sharply different outcomes. A welfare-centric government always sets the interest rate at zero to maximize the output. However, production level still falls short of the first-best at the zero interest rate because she cannot subsidize the marginal unit. A revenue-centric government charges a strictly positive interest rate when the production target is modest, creating a three-regime structure in which fiscal and environmental objectives either coexist, trade off, or become irreconcilable. Unlike conventional subsidy frameworks with the natural aim to maximize the welfare, the repayment structure of loans makes revenue-centric considerations an equally legitimate, sometimes more appropriate, policy objective. To resolve the tension between these two criteria, we introduce a two-sided green loan contract that pairs firm-side instruments with a consumer-side lump-sum charge. This richer instrument set achieves the social optimality without negative interest rates and strictly dominates both one-sided policies on welfare and fiscal grounds.

  • Strategic Resource Allocation in Dual-Channel Healthcare Systems: The Role of Telemedicine in Physician Assignment

    by Ziyang Wang, Yipei Wang, Yunzhe Qiu, Changxiao Jin

    Asynchronous telemedicine offers flexibility and accessibility to address the problem of insufficient and unevenly distributed healthcare resources. However, hospitals lack managerial guidance for resource allocation decisions across online and offline services. This study considers a hospital operating both offline and online channels, which aims to achieve economic objectives through channel capacity allocation and workload distribution between working- and after-hours. Patient medical processes are formulated as a queuing network, integrating a multi-type patient choice model. We collected related datasets from our collaborative hospital, performed simulations to validate the modeling assumptions and conducted an empirical study. Empirical results indicate that e-visits crowd out offline first-visits and substitute for offline return-visits, and the strength of the substitution effect decreases with an increase in treatment complexity. Our results suggest that increasing offline channel capacity or enhancing compensation for after-hours telemedicine services would bring a heightened preference for in-person visits during working-hours and e-visit during after-hours, which might help with economic objectives. Our analysis indicates that indiscriminate expansion of online channels may be ineffective. Instead, departmental treatment complexity should be considered, and we propose three scenario-dependent allocation strategies. This study helps hospitals make scientific decisions about developing online channels and guiding physicians’ work-load allocation.


  • Hotspot-Aware Scheduling of Virtual Machines With Overcommitment for Ultimate Utilization in Cloud Datacenters

    by Jiaxi Wu, Pavel Popov, Wenquan Yang, Andrei Gudkov, Elizaveta Ponomareva, Xinming Han, Yunzhe Qiu, Jie Song, Stepan Romanov

    We address the problem of under-utilization of resources in datacenters during cloud operations, specifically focusing on the challenge of online virtual machine (VM) scheduling. Rather than following the traditional approach of scheduling VMs based solely on their static flavors, we take into account their dynamic CPU utilization. We employ $\Gamma$ -robustness theory to manage the dynamic nature and introduce a novel variant of bin packing -(), which theoretically protects the Physical Machines (PMs) from hotspots formation within a specified probability $\alpha$ . We develop a scheduling algroithm named CloseRadiusFit and cold-start AI-based prediction algorithms for the online version of . To verify the quality of our approach towards the optimal solutions, we solve the Offline problem by designing a novel Mixed Integer Linear Programming (MILP) model and a combination of numerical upper and lower bounds. Our experimental results demonstrate that CloseRadiusFit achieves narrow gaps of 1.6% and 3.1% when compared to the lower and upper bounds, respectively. Note to Practitioners —A growing trend in the cloud industry involves overcommitting VMs on PMs. While this approach can ease the problem of low utilization of resources in datacenters, it also introduces a higher risk of hotspots due to resource contention and competition among VMs. In this work, we propose a novel method that leverages $\Gamma$ -robustness theory and introduce effective heuristics to achieve ultimate utilization of datacenter resources while ensuring desirable service quality. We validate our approach using real-world production data from Huawei Cloud, improving resource utilization by 125% over traditional flavor-based allocation methods, while maintaining the occurrence of hotspots below 5% ( $\alpha=0.05$ ). Our solution only requires VMs’ real utilization data that is typically already collected in cloud providers’ production environments. Therefore, with minimal modifications to the existing scheduling system, cloud providers can easily implement our solution and reap its benefits. Moreover, in cases of the absence of historical utilization data for VMs (cold-start), we use machine learning to predict VM utilization statistics for our approach.


  • Managing Operations of a Hog Farm Facing Volatile Markets: Inventory and Selling Strategies

    by Panos Kouvelis, Ye Liu, Yunzhe Qiu, Danko Turcic

    Problem definition: We study a dynamic finishing-stage planning problem of a pork producer who at the beginning of each week gets to see how many market-ready hogs she has available for sale and the current market prices. Then, she must decide which hogs to sell to a meatpacker and on the open market and which hogs to hold until the following week. The farmer is contracted to deliver a fixed quantity of hogs to the meatpacker each week priced according to a contractually predetermined market index. If the farmer underdelivers to the meatpacker, she pays a contractually predetermined unit penalty also linked to a market index. Biosecurity protocols prevent the farmer from buying hogs on the open market and selling them to the meatpacker. The farmer can, however, use the open market to sell hogs for prevailing market prices. Methodology/Results: We treat the problem as a dynamic, multiitem, nonstationary inventory problem with multiple sources of uncertainty. The optimal policy is a threshold policy with multiple price-dependent thresholds. The computational complexity required to evaluate the thresholds is the biggest impediment to using the optimal policy as a decision-support tool. So, we utilize an approximate dynamic programming approach that exploits the optimal policy structure and produces a sharp heuristic that is easy to implement. Managerial implications: Numerical experiments calibrated to a pork producer’s data reveal that the optimal policy with the heuristically estimated thresholds substantially improves the existing practice (around 25% on average). The success of the proposed model is attributed to recognizing the value of holding underweight hogs and effectively hedging supply uncertainty and future prices—an insight missed in the planning actions of the current practice.Supplemental Material: The online appendix is available at https://doi.org/10.1287/msom.2023.1216 .