基于代价敏感决策树的客户价值细分(英文)

由于错误分类代价差异和不同价值客户数量的不平衡分布,基于总体准确率的数据挖掘方法不能体现由于客户价值不同对分类效果带来的影响.为了解决错误分类不平衡的数据分类问题,利用代价敏感学习技术扩展现有决策树模型,将这一方法应用在客户价值细分,建立基于客户价值的错分代价矩阵,以分类代价最小化作为决策树分支的标准,建立分类的期望损失函数作为分类效果的评价标准,采用中国某银行的信用卡客户数据进行实验.实验结果表明,与传统决策树方法相比,代价敏感决策树对客户价值细分问题有更好的分类效果,可以更精确地控制代价敏感性和不同种分类错误的分布,降低总体的错误分类代价,使模型能更准确反映分类的代价,有效识别客户价值     

银行卡的替代效应及其对狭义货币需求的影响分析

本文首先研究了我国银行卡支付额对现金的替代效应,并进一步分析了银行卡支付额对狭义货币需求的影响。研究发现:随着近年来我国银行卡产业的蓬勃发展,银行卡支付总额猛增,银行卡支付额与现金流通量之间呈指数替代关系,并建立了银行卡支付额对现金替代效应的非线性模型;银行卡支付额对狭义货币需求量的影响也是非线性的,并利用非参数局部多项式拟合建立了部分线性回归模型。这两个模型在这类研究中是首次提出,并有很好的解释作用。     

The influence of financial conditions on optimal ordering and payment policies under progressive interest schemes

In many business-to-business transactions, the buyer is not required to pay immediately after the receipt of an order, but is instead allowed to postpone the payment to its suppliers for a certain period. In such a situation, the buyer can either settle the account at the end of the credit period or authorize the payment later, usually at the expense of interest that is charged by the supplier on the outstanding balance. Some payment terms, which are often referred to as trade credit contracts, contain progressive interest charges. In such cases, the supplier offers a sequence of credit periods, where the interest rate that is charged on the outstanding balance usually increases from period to period. If a buyer faces a progressive trade credit scheme, various options for settling the unpaid balance exist, where the financial impact of each option depends on the current credit interest structure and the alternative investment conditions. This paper studies the influence of different financial conditions in terms of alternative investment opportunities and credit interest structure on the optimal ordering and payment policies of a buyer on the condition that the supplier provides a progressive interest scheme. For this purpose, mathematical models are developed and analyzed.     

The Influence of Quality Inspections on the Optimal Safety Stock Level

Due to yields of less than 50% during the production of curved glass for the displays on their new cell phone series, Samsung has to deal with higher than expected production costs of several million dollars. Where there is random yield, production costs as well as holding costs can be reduced by introducing quality inspections, in which defective items are discarded before further production. To achieve the greatest cost savings, it is important to determine the optimal number and positions of these inspections across the production process which, due to several influencing parameters, is not simple. We show how the positions of inspection within a production process influence the safety stock level that is required to buffer against uncertainties due to demand and yield randomness. Our approach is the first one, combining decisions about the number and positions of inspections with inventory control strategies in a warehouse. We achieve a maximum safety stock reduction of more than 30% in our examples, which can be even larger depending on the parameter setting. For a company like Intel, reporting inventories for finished goods of nearly 1.5 billion dollars in the 2014 annual report, this allows for significant savings.     

基于ACC支付模式的供应链金融模型研究

在零售商存在资金不足和顾客需求随机情况下,以报童模型为基础,针对变质品构建了供应链金融问题和ACC支付模式（提前支付Advance、现金支付Cash和信用支付Credit）相结合的零售商订货模型,即零售商与供应商、融资机构合作条件下的模型。研究表明,零售商会随着最优订货批量和最低期望成本的增加和降低趋势,进而调整订货量,从而规避供大于求的风险,最终降低总的期望成本。同时供应商可以适当调整支付比例,让零售商来分担坏账和商品变质风险。     

基于生存分析法的顾客生命周期计量研究——以商业银行信用卡顾客为例

截至目前,国内学者对于顾客生命周期的计量问题,无论是方法研究方面还是实证研究方面都未形成统一的结论。在生物学、医学等领域较为流行的生存分析方法基础上,针对实际的顾客生存资料调研数据,采用威布尔分布法和乘积极限法来估计商业银行信用卡顾客的生命周期,得出信用卡顾客的半数生存期为30个月的结论。这种将参数与非参数相结合的计量方法为国内信用卡顾客的生命周期计量提供了一种思路。     

独立定价还是共同定价:新支付工具的定价策略

基于新支付工具发展初期市场特征构建双边市场定价模型,分析共同定价和独立定价策略下企业利润和社会福利最大化新支付工具定价。研究显示：政府较企业对新支付工具制定更低价格水平促使用户支付选择转移,特别是在共同定价策略下会有效提升市场效率,达到社会福利最大;但若政府无法做到共同定价,特别在新支付工具成本优势不明显时,会使支付工具间存在过度补贴而降低社会福利,这时企业利润最大化定价为次优选择。     

付费主体差异对信用评级结果的影响机制研究

2010年我国首家投资人付费评级机构——中债资信成立,由于我国投资人付费评级机构成立时间不长,目前国内鲜有研究投资人付费与发行人付费评级差异及产生原因的文献,并且国外相关文献主要关注评级方式对某一类债券评级结果的影响,而没有探讨两种评级方式对不同类型债券的影响之间是否存在差别。本文针对投资人付费与发行人付费评级之间的差异,基于声誉效应、竞争机制对发行人付费评级机构迎合选择进行了博弈分析,并分别以信贷资产支持证券和企业债主体的评级数据为样本,对评级差异及其产生原因进行了实证检验。研究结果表明：投资人付费评级结果显著低于发行人付费评级结果;不同付费模式下评级结果的差异受到是否为首次评级和发行人付费评级机构市场份额的影响,发行人付费评级机构对发行人的迎合是产生评级差异的主要原因;两种付费评级方式在企业债主体中的评级差异显著高于在信贷资产支持证券中的评级差异,表明付费评级方式对企业债主体的影响更大。本文较为全面地研究了我国投资人付费与发行付费评级之间的差异,使投资人、监管机构对不同付费模式下的评级结果有更加清晰的认识。     

Optimal payment time under permissible delay in payment for products with deterioration

A permissible credit period is usually allowed to a retailer to pay back the dues without paying any interest to the supplier. The retailer can pay the supplier either at the end of the credit period or later incurring interest charges on the unpaid balance for the overdue period. The retailer is expected to settle the account at a time before the end of the inventory cycle time because the payable interest rate is generally higher than the earned interest rate. A model for optimal cycle and payment times is developed here for a retailer in a deteriorating-item inventory situation where a supplier allows a specified credit period to the retailer for payment without penalty. Under these conditions, this supplier-and-retailer system is modelled as a cost minimization problem to determine the optimal payment time under various system parameters. An iterative search procedure is applied to solve the problem, and the overall findings indicate that the retailer always has an option to pay after the permissible credit period depending on unit purchase and selling price, the deterioration rate of the products and the interest rate.     

MUST‐TAKE CARDS: MERCHANT DISCOUNTS AND AVOIDED COSTS

Antitrust authorities often argue that merchants cannot reasonably turn down payment cards and therefore must accept excessively high merchant discounts. The paper attempts to shed light on this must‐take cards view from two angles. First, the paper gives some operational content to the notion of must‐take card through the avoided‐cost test or tourist test: would the merchant want to refuse a card payment when a non‐repeat customer with enough cash in her pocket is about to pay at the cash register? It analyzes its relevance as an indicator of excessive interchange fees. Second, it identifies four key sources of potential social biases in the payment card systems' determination of interchange fees and compares the industry and social optima both in the short term (fixed number of issuers) and the long term (in which issuer offerings and entry respond to profitability).     

中国居民性别差异对消费信贷盈利贡献的实证测度——来自商业银行信用卡业务的证据

为了测度性别差异对消费信贷盈利贡献的影响机制，本文基于商业银行信用卡业务103455个账户及其过去一年的交易记录，构建了性别差异对消费信贷盈利贡献的中介效应模型。结果显示：在综合效应方面，女性对消费信贷的盈利贡献显著高于男性。在中介效应和直接效应方面，性别差异通过影响交易频率对消费信贷盈利贡献有显著影响，贷款规模和风险水平对盈利贡献有负向完全中介效应，交易金额、忠诚度、最后一次交易距离目前的时间对盈利贡献无显著影响。说明消费金融机构应改变规模驱动的粗放发展路径，转向提升交易频率的客户经营策略。我国应大力发展女性经济及消费生态，提升整体贡献水平。     

Money and Credit Redux

We analyze money and credit as competing payment instruments in decentralized exchange. In natural environments, we show the economy does not need both: if credit is easy, money is irrelevant; if credit is tight, money is essential, but credit becomes irrelevant. Changes in credit conditions are neutral because real balances respond endogenously to keep total liquidity constant. This is true for both exogenous and endogenous debt limits and policy limits, secured and unsecured lending, and general pricing mechanisms. While we show how to overturn some of these results, the benchmark model suggests credit might matter less than people think.     

Taxes and the choice between risky and risk-free debt: on the neutrality of credit default taxation

Shareholders can decide if their corporation issues risky or risk-free debt. We identify tax systems in which the choice between risky and risk-free debt is not distorted by taxes. These credit default neutral tax systems make it possible to make capital structure decisions and firm valuations neglecting credit default risk, even after taxes. Thus credit default neutrality is a characteristic of a tax system that helps to reduce planning costs. Moreover, credit default neutrality is a necessary condition for financial neutrality of taxation. We find one class of credit default neutral taxes that preserves and another class that modifies the expected tax distribution between creditors and debtor firm. Finally, we show the influence of personal taxation on credit default neutrality.

考虑纵向持股的现代农业供应链运营决策与协调优化研究

针对农业数字化技术的高成本破解难题及供应链企业间鲁棒性较弱的现象,将纵向持股策略引入到现代农业供应链中,分别构建了公司对数字化技术投资与否、农场对该投资持股与否等不同情景组合下的博弈模型,研究股权合作下现代农业供应链运营决策及协调优化问题。研究发现：投入数字化技术能有效提高农产品品质,实现农场丰产增收,但公司单独对农产品投资该技术使农场产生“搭便车”现象;农场对公司投资数字化技术在一定比例范围内持股,农产品品质得到持续改善,供应链企业决策和绩效实现Pareto改进,同时持股行为使主导者公司获得更多的利润增量,消除了农场“搭便车”现象;结合公司和农场实际运营场景,在股权合作策略下采用交互式Rubinstein讨价还价模型实现了农业供应链的协调和优化。最后,通过数值分析对主要结论进行了充分验证,并探讨了持股比例和品质增量成本系数同时变动对供应链运营决策和绩效的影响。     

延期支付条件下的最优付款时间确定：考虑现金折扣情形

现有延期支付条件下零售商最优付款时间的研究文献均假设,零售商付款时间可以在延期支付期限时点或补货周期之内任一时点,但这一关键假设却并不符合商业惯例,也没有考虑实践中卖方广泛使用现金折扣以刺激买方提前还款的情形。为此,通过对已有模型进行了修正和补充后,给出了零售商在供应商给定延期付款、现金折扣政策下最优付款时间的判定方法。     

Replacement Decisions for Potentially Hazardous Substances

As public awareness of environmental hazards increases, a growing concern for corporations is the potential negative environmental impact of their products and the chemicals these products contain. In this study, we analyze the optimal decisions of a firm when a substance within its product is identified as potentially hazardous. Although the substance is not currently regulated, regulation may occur in the future. Therefore, the firm must devise a strategy for the development and implementation of a replacement substance. In an environment where replacement costs can be millions of dollars, regulation is uncertain, and both consumer and non‐governmental organization pressures exist, a carefully developed plan that balances costs and risks is critical for a firm. Our results demonstrate that as long as a threat of regulation exists, a firm should always dedicate resources toward developing a replacement substance. However, it is not always optimal for a firm to implement a developed replacement. Regarding competitive dynamics, we find that competition between firms can offset a low chance of a shift in consumer perception about a substance and compel firms to replace; however, competition can lead to inefficient outcomes in which firms incur avoidable costs to implement ahead of potential regulation.     

基于银行授信额度的Max-NPV项目调度问题研究

本文对基于银行授信额度的Max-NPV项目调度问题进行研究.首先界定研究的假设条件并对文中所用到的符号进行定义;随后采用基于事件的研究方法构建了问题的整数规划优化模型;鉴于问题的NP-hard属性,设计了双层模拟退火搜索循环嵌套的启发式求解算法;最后对一个算例进行了求解分析,并讨论了银行授信额度及客户支付比例对项目收益的影响.结果表明:随着银行授信额度的提高,承包商安排项目进度自由度的增加使得其净现值得到改善,在支付比例较低的条件下项目完成时间也随之提前;当客户支付比例提高时,项目净现值单调上升,而项目完成时间呈先提前后延迟的变化趋势.     

基于计划行为理论的信用卡使用意向分析及其营销策略研究

在计划行为理论的基础上,使用结构方程模型等数理统计方法,探析了消费者信用卡使用意向的影响因素,建构了消费者信用卡使用意向模型。研究表明,计划行为理论中信用卡态度和知觉行为控制变量对消费者信用卡使用意向有重要影响。此外,研究还进一步将消费者信用卡态度区分为便利性态度、金钱与信用态度,实证研究显示便利性态度对信用卡使用意向有重要影响,并且这种区分是合理的和必要的。     

THE TRADE DISCOUNT DECISION: A MARKOV CHAIN APPROACH*

Economic theory indicates that higher returns are required from investments that have higher risk. Two major reasons for offering trade discounts are to stimulate sales and to speed up cash receipts. Both sales increases and the earlier receipt of cash impact the risk/return characteristics of a firm. This study uses a Markov chain model to demonstrate how alternative trade credit policies impact the risk and required returns of a firm. The amount of risk and return per credit sale is calculated, and the coefficient of variation is used as a risk/return measure per credit sale. The study concludes that trade discounts can have a favorable impact on the risk/return characteristics of a firm, even in the absence of increased sales volume. Other discount decision factors which are either directly or indirectly determined in this paper include: (1) the average number of periods for which an account is outstanding, (2) the probability of collection and bad-debt losses over the average account period, (3) the average speed of payment, and (4) the average amount of cash tied up in accounts receivable.     

Financing Liabilities Under Superfund

One of the more critical environmental risk areas involves the financing of cleanup costs associated with past improper disposal of hazardous wastes. These costs will run into the tens of billions of dollars. The federal Superfund program and related state programs are the driving forces for cleaning up hazardous waste. Under Superfund, the government collects taxes and coordinates cleanup strategies, and searches for potentially responsible parties (generators, disposers, transporters) to assist in financing the cleanup. To lessen the financial impact, responsible parties are looking to their liability insurers to provide coverage under old general liability insurance policies. Insurance companies contend that Superfund liabilities are not covered under liability policies. The paper examines the various financing methods and liabilities produced by Superfund. Particular emphasis is placed on developments in the courts which are resulting in a significant shift in the financial responsibilities to the insurance industry. Broad estimates of the financial impact of Superfund liabilities are developed. Finally, several public policy issues, which are raised by the topics exmained in this paper, are discussed.