The Financial Markets of Roman Egypt: Risk and Return.Paul V.Kelly, (Liverpool: Liverpool University Press, 2023. pp. vii + 221. 36 figs. 15 tabs. ISBN: 9781802078336, Hbk £76)
本书利用大量纸莎草文献数据，量化分析了罗马埃及的住房、土地和信贷市场，计算了不同投资的风险调整后回报，发现土地和抵押贷款市场存在理性参与者，而住房和无担保贷款市场则因信息不足、交易成本高等因素表现非理性。
Paul V. Kelly aims at analysing the financial markets of Roman Egypt, that is, the housing, the land, and the credit markets. Relying heavily on an unpublished PhD thesis, the book is roughly divided into four sections. After the introduction, a first chapter provides the reader with a basic description of the currency and price systems. Chapters 3–5 describe the different markets, with their simulated returns on capital. Chapter 6 focuses on the role of women. Finally, chapter 7 offers a risk–return comparative assessment, leading to a general conclusion dealing with economic rationality in Roman Egypt in chapter 8. Undoubtedly, this book embodies an ambitious application of quantitative data analysis in exploring market dynamics in the ancient world. Different authors have used the vast papyrological corpus to estimate rents, housing, and land prices, as well as interest rates, in Ptolemaic and Roman Egypt. However, none had compiled such a substantial database and subjected it to rigorous statistical analysis. Indeed, Kelly used 2941 documents dealing with 4367 transactions ranging between 1 and 350 CE (p. 16), the price charts incorporate 408 data points, and 1008 loans are tabulated (pp. 26–7). This stands as a significant accomplishment. Kelly offers a comprehensive analysis of these investments’ risk-adjusted returns. His findings imply that cash loans, even with the frequent use of 12 per cent as interest rate, returned almost nothing to the lenders once default risks were accounted for. Investments in land and mortgages offered around 7–8 per cent, and loans in kind almost 25 per cent (pp. 99, 100, 160–1, and 166 notably). Compared with twenty-first century asset classes and their risk–return performance, these findings lead Kelly to posit that some of these markets involved rational players – land and mortgage loans – while these conditions were not met by the housing market, plagued by insufficient information, high frictional costs, and a high level of intra-familial transactions. The same would apply to the unsecured loan markets, the low returns being explained by reciprocity considerations, and the loans in kind, hampered by high variability in outcomes (findings summarized pp. 190–5). While these conclusions carry potentially far-reaching implications, the reader remains somehow frustrated that most of the data and methods remain buried in an unpublished dissertation. That ‘this methodology is described in my thesis’ (p. 15) does not help much. An open database, such as the one provided by Kyle Harper in support of ‘People, plagues, and prices in the Roman world: the evidence from Egypt’ (Journal of Economic History 76, 803–39), would have been very useful. When Kelly states that 18 942 documentary papyri could belong to the 1–350 CE range, the sentence implies loose dating (‘could fall’, p. 15), which could create significant issues when tabulating inflation-adjusted prices, a critical step in the author's method (appendix, pp. 198–9). Likewise, the earliest seven land prices exhibit remarkable divergences, questioning the concept of a single average price during that period (figure 9, p. 27). Maybe regional differences or specific types of transactions are at stake. While the author uses 1008 loans for chart 13 (p. 27), he speaks in the chapter that follows of 1999 loans (p. 62). The reader will suspect some of these loans are not dated, hence the discrepancy, but this is not explained. There is no explicit mention that Monte Carlo simulations are used to obtain the stochastic results. An indirect reference is encountered once (n. 21, p. 7), leading to a definition in the appendix (p. 204), which does not specify that Monte Carlo powers the 1000 simulations applied to a group of 105 individual leases for instance (p. 99). Some of the conclusions offered by Kelly remain debatable. He characterizes the rental housing market as irrational because of, among other factors, ‘the relatively high level of intra-familial transactions’ (p. 192). However, these transactions represent 9 per cent of all trades (pp. 50, 61, 183) and belong to the housing sales market and not to rentals. In any case, would 9 per cent derail an entire market, especially when many of these family transactions did not fit within the reciprocity framework (p. 4)? Similarly, the author's rejection of cash loans as irrational relies on the assumption that all defaults had a 0-recovery rate (pp. 136, 159–61). Some of the situations documented by the papyrological corpus are more complex, as exemplified by the Kronion family loans around 138 CE (p. 186). New loans were used to repay older ones, and when the family situation worsened, collateral land, personal bondage, and prodomatic leases were used to address their growing indebtedness. Full defaults, in a rural context where the loss of one's reputation was irrecoverable, was avoided at all costs, including offering oneself as dependent labourer. This would improve the lenders’ prospects very significantly, challenging the idea they were not profit-driven. Finally, lenders were overwhelmingly city dwellers, while borrowers lived in villages (p. 129, table 11), weakening the notion that some form of reciprocity might explain inferior financial outcomes. This said, the author engaged very convincingly with the underlying corpus, and several of his conclusions challenge decisively some well-established consensual views – such as the idea Egyptians prices would have been stable during most of the third century until 274 CE, or that 12 per cent represented universally fixed loan interest rates during that period. He demonstrates as well that the grain market's behaviour was sensitive to state interventions and that the impact of the so-called Antonine Plague was probably quite limited. Overall, the book offers a comprehensive overview of the various markets of Roman Egypt and their different players, a rare achievement for the ancient world. The use of ‘financial markets’ rather than simply ‘markets’ in the title could be questioned at the end since the concept of tradable financial instruments remains mostly anachronistic for that period. Kelly's utilization of statistics highlights the potential that quantitative techniques offer to the field of ancient economic history. The book is a must-read, and we should hope that the underlying database will be made available.