Natural Selection in Markets

对于企业利润最大化这一假设，最常见的理论辩护是，否则它们将被赶出市场。这个想法是，只有那些无论出于何种原因成功实现利润最大化的公司才能够吸引或积累足够的资本来生存和发展。虽然这种直觉已经存在了很长时间（见阿尔奇安（1950），弗里德曼（1953）和温特（1964，1971）），但它还没有在一般均衡背景下进行过研究。这一论点也遭到了批评。特别是，Koopmans（1957）认为，引用外部动态过程来支持关键行为假设的有效性并不是一种令人满意的方法。 在导致这一提议的工作中，我们已经在一个简单的一般均衡模型中检验了这一假设，在这个模型中，企业从留存收益中为当前业务提供资金，而不进入外部金融资本市场。这种动态，就像温特所讨论的那样，驱动着企业经营的规模。我们发现，在所有获得相同技术的公司中，只有那些最接近利润最大化的公司才能生存。但库普曼斯仍然是对的。选择利润最大化者的倾向并不会产生对总生产者效率的相应选择。在没有金融资本市场的情况下，留存收益动态无法在企业之间重新分配资金，从而实现生产者有效的结果。资本市场的加入“解决”了这个问题，因为如果所有投资者都有理性预期，并且市场是动态完全的，那么均衡是有效的。但这引出了一个问题。更有趣的是，如果不是所有投资者都有理性预期，会发生什么。在这种情况下，结果可能比没有金融资本市场更糟。 我们的研究计划可以说是认真对待库普曼的担忧。我们感兴趣的是，当企业不一定能实现利润最大化，投资者也不一定是理性的，但经济自然选择的力量在起作用时，会发生什么。均衡序列显示出什么特征？渐近性质是什么？均衡序列或极限均衡是否有最优性性质？我们知道，利润最大化者在某些情况下被选择，而在其他情况下则不是，但合适的决策规则是否可以被一般地描述？ 当利润是随机的，公司不能通过套利来估值时，这些问题特别有趣。在这种情况下，不清楚公司的目标是什么。我们的初步调查表明，即使资本家想要的是期望利润最大化，也没有选择的倾向。相反，我们推测，将有选择的公司采取行动，以最大限度地提高其预期增长率的留存收益。这与预期利润最大化不一致，但与关于大公司目标的流行评论一致。

The most common theoretical defense of the hypothesis that firms maximize profits is that otherwise they will be driven out of the market. The idea is that only those firms which for whatever reason succeed in maximizing profits will able to attract or accumulate enough capital to survive and grow. Although this intuition has been around for a long time (see Alchian (1950), Friedman (1953) and Winter (1964, 1971)) it has not been investigated in a general equilibrium setting. The argument also has its critics. In particular, Koopmans (1957) argued that referencing an external dynamic process to support the validity of a key behavioral assumption is not really a satisfactory way to proceed. In the work leading up to this proposal we have examined this hypothesis in a simple general equilibrium model in which firms fund current operations from retained earnings, without access to external financial capital markets. This dynamic, much like that discussed by Winter, drives the scale of firm operation. We found that among all firms having access to the same technology only those that come closest to maximizing profits survive. But Koopmans is nonetheless right. The tendency to select for profit maximizers does not generate a corresponding selection for aggregate producer efficiency. In the absence of markets for financial capital, the retained earnings dynamic is not able to reallocate funds across firms so as to achieve producer efficient outcomes. The addition of capital markets `solves` the problem in that if all investors have rational expectations, and markets are dynamically complete, then the equilibrium is efficient. But this begs the question. More interesting is what happens if not all investors have rational expectations. In this case the result can be even worse than without financial capital markets. Our research plan can be characterized as taking Koopmans' concern seriously. We are interested in what does happen when firms do not necessarily maximize profits, and investors are not necessarily rational, but the forces of economic natural selection are at work. What characteristics does the sequence of equilibria display? What are the asymptotic properties? Are there any optimality properties for either the sequence of equilibria or for the limit equilibria? We know that profit maximizers are selected for in some settings and not in others, but can the fit decision rules be characterized generally? These questions are particularly interesting when profits are random and firms cannot be valued through arbitrage. In this case it is not clear what objective to attribute to a firm. Our preliminary investigation suggest that there is no tendency to select for expected profit maximization even when that is what capitalists want. Instead we conjecture that there will be selection for firms that act so as to maximize the expected growth rate of their retained earnings. This is not consistent with expected profit maximization, but it is consistent with popular commentary about the objectives of large corporations.