How central banks cope with the international finance constraint in a fixed-exchange-rate regime is an old debate, with the key tradeoffs typically formulated as Mundell’s trilemma (Klein and Shambaugh 2013). This debate has gained new interest with the recent focus on balance sheet policies of central banks, particularly for Asian countries (Aizenman, et al. 2009, BIS 2012). To gain insight into today’s balance-sheets issues, this column looks to the past – specifically how the Banque de France used its domestic portfolio (rather than exchange market intervention) to maintain a stable domestic discount rate during the classical gold standard period (1880-1914) despite varying international rates (Bazot et al. 2014). In short:
- When the Bank of England – the leader of the gold standard – increased its discount rate, the Banque de France expanded its domestic portfolio in order to stabilize the interest rate in the French money market (which had increased with the English rate).
- To conduct such a sterilisation policy required having a large gold stock and, when its gold reserves fell too much, the Banque de France had to increase its discount rate.
Central banks cannot forever escape the constraint of international finance.
The gold standard and the rules of the games
The textbook account of the classical gold standard emphasizes the crucial role of the ‘rules of the game’ in international adjustment. In a deficit country, the central bank would use discount-rate policy, raising its discount rate and tightening credit. In the surplus country, the central bank would lower its discount rate and expand credit. As a consequence, two components of the ‘rules of the game’ should be observed in central banks’ practices:
the central bank should raise its discount rate when its gold reserves decrease, and
central banks should not sterilize – that is, expand (or reduce) – domestic assets to offset outflows (or inflows) of gold.
The adherence to the rules implies that domestic assets of central banks should be positively correlated with international assets and negatively correlated with the domestic interest rate. As in the Mundell trilemma, the only solution in theory to breaking the rules would have been capital controls.
The Banque de France, one of the main pillars of the classical gold standard along with the Bank of England and the Reichsbank, continuously violated the two components of the rules of the games. First, the official discount rate remained flat (there were only 30 changes in the French official rate of discount compared to 194 in the Bank of England). Second, the discount portfolio of the Banque de France varied negatively with gold flows systematically between 1880 and 1913. Only a soft form of capital controls was imposed (a gold premium on gold sales) and it was abolished in 1900.
International shocks and the domestic portfolio of the central bank
We offer new evidence on these two deviations from the rules of the game. We also highlight the importance of the Banque de France’s sterilization policy (through expansion of its domestic portfolio), a fact which is much less known and documented than the stability of the discount rate – even though the latter depends on the former. More specifically, we show that the Banque de France’s domestic credit (discounts and advances) correlates negatively with gold flows because it correlates positively with the Bank of England’s discount rate (Figure 1). We estimate a VAR (Figure 2) and find that an increase in the English discount rate led to a fall in the gold reserves of the Banque de France but pushed up its domestic portfolio such that gold outflows were almost completely sterilized. Results show that domestic adjustment to an international shock took place in a period shorter than a year. Conversely, such an international shock had no significant effect on the French business cycle (as proxied by imports and railway revenues). The extent of sterilization of international shocks by the central bank explains these findings; movements in the domestic portfolio of the central bank eliminated the effect of international shocks on the domestic interest rate and the business cycle.
Figure 1. Bank of England's rate and Banque de France's portfolio
Figure 2. Impact of a change in Bank of England's rate on French central bank and economy
In a world of integrated financial markets and British leadership, the French money market interest rate was influenced by foreign rates (especially the British). A rise in the Bank of England rate decreased the spread between the French money market rate and the Banque de France official discount rate (CF VAR estimations in Figures 2 and 3). Facing a higher demand for rediscounting, the Banque could either increase its discount-rate, or conversely, increase its discount portfolio to expand domestic credit, backed by large gold reserves. The increase in its discounts brought the spread back to its previous level and kept the official discount rate stable (Figure 2). This is the crucial mechanism identified in this paper.
Figure 3. Money market rate in Paris and central banks' rates
Discount rate’s changes as regime shifts
Our elucidation of the adjustments that were necessary to keep the discount rate stable also offers a good explanation of the few changes in the Banque rate that did occur. The Banque discount rate was changed only in extreme cases when the previously highlighted mechanisms could not work fully. Thus the response of the Banque rate to key economic variables was highly nonlinear. We account for this nonlinearity in a regime switching model in order to avoid the pitfalls of previous studies, finding that the Banque de France’s decision to change its discount rate was determined by the Bank of England rate, the gold stock, and the deviation of the exchange rate from the gold points (Bazot et al. 2014). Figure 4 shows that the probability of being in regime 2 from the Markov switching model (in which the gold stock, the exchange rate, and the British rate are significant) matches the changes in the Banque de France discount rate. Our interpretation in terms of regime shifts makes clear that the changes were infrequent because, in normal times, the Banque de France used all possible means to reduce the probability that the variations of the Bank of England’s rate and of the exchange rate affect its decision to move its discount rate. Still, the Banque de France could not forever defend the stability of its discount rate and in extreme cases had to cope with the rules of the game.
Figure 4. Changes in the rate of the Banque de France and regime shifts
Two polar cases: France and England
A better understanding of the policy of the Banque de France leads to a better understanding of the classical gold standard. It is commonly argued that the Bank of England and the Banque de France were the two extremes in the spectrum of the monetary regime. These opposite positions in policy are reflected in the liquidity ratios (metallic reserves divided by notes in circulation) of these countries – France had the higher ratio whereas England had the lower (Morys 2013).
While the objectives and interventions of the Bank of England under the gold standard have been studied extensively, we provide a complete interpretation and account of the policy of the Banque de France. An examination of the balance sheet policy of the Banque de France – which was the opposite of the Bank of England’s flexible rate policy – can provide new insights for the study of central banks which fell in between the two poles on the continuum. Smaller central banks under the gold standard used exchange market intervention to ease the international constraint (Esteves et al. 2009, Jobst 2009). But none could rely – as the Banque de France did – solely on its domestic portfolio.
Overall, our approach adds new dimensions to the traditional view of the gold standard that assumes that central banks' discount rates and gold devices (disguised capital controls) were the most important monetary policy instruments. Focusing solely on interest rates to study domestic policy objectives and international adjustments might be misleading.[R1]
Aizenman, J, M D Chinn, and H Ito (2009), “Surfing the Waves of Globalisation: Asia and Financial Globalisation in the Context of the Trilemma”, Asian Development Bank Working Papers No. 180
Bank of International Settlements (2012), Are central bank balance sheets in Asia too large? BIS Papers No 66.
Bazot, G, M D Bordo, and E Monnet (2014), “The Price of Stability: The balance sheet policy of the Banque de France and the Gold Standard (1880-1914)”; NBER Working Paper No. 20554.
Bordo, M, O F Humpage, and A J Schwartz (2012), “Foreign-exchange intervention and the fundamental trilemma of international finance: Notes for currency wars”, VoxEU.org, June 2012
Esteves, R P, J Reis, and F Ferramosca (2009), "Market Integration in the Golden Periphery. The Lisbon/London Exchange, 1854-1891", Explorations in Economic History, vol. 46(3), p. 324-345, July.
Jobst, C (2009), “Market Leader: The Austro-Hungarian Bank and the making of foreign exchange interventions, 1896-1913”, European Review of Economic History, 13, pp. 287-318.
Klein, M W and J C Shambaugh (2013), “Is there a dilemma with the Trilemma?”, VoxEU.org, 27 September 2013
Morys, M (2013), “Discount rate policy under the Classical Gold Standard: core versus periphery (1870s -1914)”, Explorations in Economic History, 50, pp.205-226.