This paper uses recent advances in Bayesian multiple time series econometrics to study the link between merchant bank trade financing and economic conditions in Britain in 1880-1913. Recent research in macroeconomics has highlighted the importance of trade credit for explaining fluctuations in international trade and macroeconomic conditions. While there are several qualitative accounts of how trade was financed in the pre-WW1 years, the quantitative literature studying the link between trade finance and the economy in this era is very limited. In order to study this link, the paper presents a new dataset on merchant bank acceptances, which were effectively means of insuring credit payments between two parties. London merchant banks played a dominant role in the financing of international trade before WW1, and there are several reasons why we might expect that there were constraints on the amount of trade credit they issued. While using multiple time series econometrics is a commonly used framework to study the macroeconomic effects of financial shocks, several features of the underlying data make the inferences drawn from a simple structural vector autoregression (VAR) questionable, and necessitate the use of more advanced methods. First, given that there were several large fluctuations in the acceptance market, such as the Baring crisis, along with a growing internationalisation of acceptance credit, it is reasonable to assume that shocks in acceptances had varying effects in different time-periods, and that the model should take into account this time-varying relationship. Second, because no quarterly GDP data exists for the period being studied, alternative means of approximating economic conditions are needed. Summarizing a large number of macroeconomic series to their common components, and incorporating these into a VAR has become a useful way of dealing with an absence of high frequency GDP figures and reducing omitted variable bias. Therefore, this paper employs a factor augmented vector autoregression with time-varying parameters, which is able to handle several of the complexities in the underlying data. I find that shocks in acceptances impacted macroeconomic conditions and exports, but that the effect decreased over time. I hypothesise that this is because a growing amount of trade financed through London was not linked to British trade. Furthermore, I highlight a potential link between the London money market and trade credit.