Development Banking

Outline

• Define the problem;
• History of development banking;
• Analytical model;
• Case studies of France and Mexico.

Financial institution, that satisfy following properties:

• Provides long-term finance;
• Sponsored by the national government;
• Finance primarily new sectors and enterprises.

• Risk of the new industries;
• Long-term risk.

• Gershenkron (1952), Cameron (1953), Diamond (1957) identify 19th century industialization with rapid growth of large financial institutions;
• Acemoglu and Zilibotti (1996) links the degree of market incompleteness to capital accumulation;
• Dewatripont and Maskin (1995) introduce a role of coordinating agencies to overcome free-rider problem;

• The oldest government-sponsored institution for industrial development was Societe General pour Favoriser I'Industrie National in Netherlands (1822);
• Several decades later France established several institutions such as Credit Foncier, the Comptoir d'Escompte, and the Credit Mobilier. Particularly, Credit Mobilier funded Continental European railway which allowed it to acquire substantial expertise. Following example, similar institutions were founded in Japan and India;
• After World War I, reconstruction gave a new push to the development of government institutions. Most notably, these institutions showed greater success in diffusion of expertise because it was owned by several financial intermediaries that can only provide supplementary finance;
• After Second World War, another surge of development bank started to emerge. These institutions were primarily state-owned and had obligation to participate in co-financing arrangements.

• There is one project, that yields return \( R\) with probability \( p\) if completed;

• Co-financing requires to acquire expertise for all participants investing into the project;

• Probability of success \( p\) depends on monitoring from all the banks involved in financing (\(p=m_1+m_2\)).

• Each project requires total sink cost \( K\). In two bank cases, each of them provide \( \frac{K}{2}\);

• Monitoring for each bank is denoted by \( m=\{\underline{m},\overline{m}\}\);

• Monitoring assumed to incur cost \(d\) to the bank, with \(d(\underline{m})=0\) and \(d(\overline{m})=c(\mu)\), where \(\mu\leqslant 1\) - share of transmitted expertise to the bank. If bank itself acquired expertise, i.e. \(\mu=1\), cost is \(c(1)=0\). Assume \(c'<0\);

• Acquisition of expertise requires to provide non-monetary effort from bank manager. I.e. to acquire expertise with probability \(e\), manager must incur \(g(e)\), where \(g'(e)>0,\, g''(e)>0\). Probability to start the project in two bank case is thus \(e=1-(1-e_1)(1-e_2)\). For calculations author assumes \(g(e)=\gamma e^2\), \(\gamma\gg 0\);

• Expertise itself is source of proprietary knowledge to the bank manager. He values it at \(B(1-\mu)\);

• Expertise cannot be described ex-ante, banks cannot bargain over \(e\) or \(\mu\) in advance.

\[ U_i=(1-(1-e_i)(1-e_j))p(m_i,m_j)\frac{R}{2}+e_i(1-e_j)B(1-\mu_i)-\frac{K}{2}-c(\mu_i)-g(e_i) \]

• Assume expertise is non-rival good that is automatically and costlessly transmit to the entire financial system. Then symmetric Nash equilibrium will be \(\hat{e}_1=\hat{e}_2=\frac{\overline{m}R}{2\gamma+\overline{m}R}\);

• Compared to social optimum \(e^\ast=\frac{\overline{m}R}{\gamma+\overline{m}R}>\hat{e}\), it can be seen there is underinvestment of expertise;

• This results arise due to strategic substitability of the \(e_1\) and \(e_2\).

• Consider manager of the bank that acquired expertise can choose optimal level of \(\mu\) to be transmitted;

• Ex-post social optimum of \(\mu^\ast\) will be achieved by optimizing:

\[ \max_{\mu} (\overline{m}+m_2)R+B(1-\mu)-K-c(\mu)\\ \text{s.t.}\begin{cases} &m_2=\overline{m},\, \text{iff } (\overline{m}+\overline{m})\frac{R}{2}\geqslant c(\mu),\\ &m_2=\underline{m}, \, \text{otherwise}. \end{cases} \]

• \(\mu=max(\mu^\ast, \overline{\mu})\)

• Ex-ante \(e^\ast\) is then given by:

\[ e^\ast=arg\max_e (1-(1-e)^2)(2\overline{m}R+B(1-\mu^\ast))-g(e)=\frac{\overline{m}R+\frac{B}{2}(1-\mu^\ast)}{\gamma+\overline{m}R+\frac{B}{2}(1-\mu^\ast)}. \]

• Similarly to the social optimum, optimal \(\mu\) will be chosen by bank 1:

\[ \max_{\mu} (\overline{m}+m_2)R+B(1-\mu)-K\\ \text{s.t.}\, m_2=\overline{m},\, \text{iff } (\overline{m}+\overline{m})\frac{R}{2}\geqslant c(\mu); \]

• Ex-ante effort \(e_i\) is found by maximizing:

\[ \max_{e_i}(1-(1-e_i)(1-e_j))2\overline{m}\frac{R}{2}+e_i(1-e_j)B(1-\overline{\mu})-g(e_i); \]

• Symmetric Nash equilibrium will be \(\hat{e}=\frac{\overline{m}R+\frac{B}{2}(1-\overline{\mu})}{2\gamma+\overline{m}R+\frac{B}{2}(1-\overline{\mu})}\)

• In decentralized system, there is potentially less information transmission of expertise;

• Assuming information transmission is the same, there is underinvestment in expertise;

• However, undertransmission of expertise mitigate underinvestment. Risk of concealing expertise may lead managers to overinvest (depend of \(B\)).

• Suppose government subsidize the bank to finance certain industries;

• Condition to invest into new industry will be:

\[ \overline{p}\frac{R_0}{2}-\frac{K_0-\Delta K}{2}+B_0 < (2\overline{m})\frac{R}{2}-\frac{K-\Delta K}{2}+B(1-\mu)-g(\hat{e}); \]

• Policy implications: condition government to support certain industries.

• What if bank can undertake project alone, while taking the risk \(\sigma< \infty\)?

• Co-financing still occur iff:

\[ \Pi^{nc}=R-K+B-\sigma<\Pi^{c}=R-K+B(1-\mu); \]

• Policy implications: make government support conditional upon co-financing.

• Suppose bank 1 owns \((1-\alpha)\) share in bank 2;

• Profit of the bank 2 will be:

\[ \Pi=\alpha(2\overline{m})\frac{R}{2}-\alpha\frac{K}{2}+B(1-\overline{\overline{\mu}}); \]

• Contrary to intuition transmission \(\overline{\overline{\mu}}<\overline{\mu}\);

• Co-ownership will also negatively affect incentive to acquire expertise.

• Credit National acquire substantially more expertise than Nacional Financiera;

• Both were established by respective governments (and supported by them) to provide long term finance, among other things;

• Credit National had three requirements for government support:

  1. It has to provide long term loans;
  2. It gives priorities to the privileged sectors;
  3. it has to finance private owned companies.

• Credit National had also very dispersed ownership. The government encouraged co-financing known as "Fonds Mobilisables";

• On other hand, Nacional Financiera were far less specialized and had to perform other functions;

• Framework of co-financing the new industries was built upon acquisition and transmission of expertise;

• Competitive Nash is generally underinvest in expertise compared to social optimum;

• Under laissez-faire system, undertransmission of expertise occur that can alleviate underinvestment problem;

• Joint ownership may have adverse effect on transmission of expertise.