Saturday, 27 April 2013

Change at the IMF: interconnectedness, financial fragility and global banks

I attended recently a workshop organized by Cornel Ban and Kevin Gallagher, with support from Governance, at Boston University on how the crisis has changed the IMF.

My presentation explored the way in which the IMF has grappled with financial innovation since the crisis, both theoretically and in its policy advice.

This is important, it argues, because the global economic crisis has brought an important shift in the IMF’s understanding of crisis, its triggers and its actors. Its research on macro-financial linkages now identifies large capital inflows as the main conduit for the transmission of global shocks, in contrast to previous concerns with current account dynamics; and transnational banks as the key carriers of capital flows across borders. With this, the IMF has included private financial institutions in its analytics of crisis, previously focused on governments (fiscal policy), central banks (monetary policy) and trade union/state-owned companies (structural reform). 

The IMF uses interconnectedness as an analytical lens to capture the links between links financial institutions and markets. This is in line with the rapidly growing consensus on macroprudential policies, a consensus that views systemic risk arising endogenously from a combination of financial cyclicality and interconnectedness. What is remarkable though is the departure from financial programming/inflation targeting (that is, both the economics of crisis and economics of stability at the IMF) treatment of financial institutions as passive intermediaries in a process largely determined by the central banks' monetary policy decisions. In place of previously invisible (analytically) financial institutions the IMF now speaks of global banks acting as key nodes in global financial architectures, with complex business models reliant on cross-border funding and investment patterns; trading and market making activity. More remarkable still, IMF research describes transnational banks as important political economy actors, arbitrageurs across uneven regulatory, tax and yield landscapes. This analysis of the complexity of modern global finance echoes Ilene Grabel's poignant observation that 'your grandfather's IMF' has made way for a research and policy agenda of 'productive incoherence' pregnant with possibilities for developing countries. Indeed, in contrast with its previous insistence on unchecked financial globalization, the IMF now affirms that countries may want to limit participation in cross-border financial networks by actively changing the mechanisms of interconnectedness through which transnational banks generate or carry systemic risks in(to) the national economy.

My paper explores how this treatment of interconnectedness as generator of systemic risk guides the IMF's policy advice in developing countries. To do so, it first proposes an analytical framework for identifying the mechanisms through which transnational banks construct cross-border interconnectedness, defined as fragile complementarities. This analytical framework brings together research that the IMF has conducted broadly within the theme of macro-financial linkages/capital controls/macro-prudential policies. It then asks how the IMF operationalized its concerns with fragile complementarities in two settings: the multilateral negotiations of the Vienna Initiative to preserve transnational banks’ exposure to Central and Eastern European countries after Lehman’s collapse; and  bilateral relationships between the Fund and four Latin American countries that threaten to ‘go Brazilian’ in early 2013 and impose or tighten capital controls (Mexico, Peru, Chile, Colombia). 

I modify the concept of institutional complementarities developed in the varieties of capitalism literature. According to this, two sets of institutions are complementary when their interactions increase returns for both, and improve aggregate economic performance. When applied to macro-financial relationships, in aggregate such complementarities generate fragility as an externality: individual gains from transactions increase systemic risk in the aggregate. Such fragile institutional complementarities can be traced back to four basic relationships:

1.     transnational banks – domestic banks (across interbank money and currency markets)
2.    transnational banks – central bank policies (across currency markets)
3.     transnational banks – non-resident investors (through carry-trade strategies)
4.    transnational banks – non-bank financial intermediaries (shadow banking), which I wont discuss here because the paper focuses on IMF advice in developing countries.

1. Transnational banks - domestic banking system.  

Before 2008,  transnational banks used to allocate liquidity and capital across subsidiaries freely or to lend to domestic banks in developing countries. Host countries chose to view the positive effects: this allowed the domestic banking system to overcome the constraints of low savings rate, improved risk management and efficiency of capital allocation because transnational banks were typically portrayed as better equipped to manage risk. Few countries worried about credit booms, and fewer still imposed capital controls. But the crisis made apparent the systemic risks embedded in cross-border banking flows and that not only banking systems, but also regulators had become interconnected. When parent banks ran into trouble because of exposures to complex financial products in the US, host regulators realized that subsidiaries may stop receiving funding through internal capital markets of the banking group. Worst still, no political mechanism was in place for home and host regulators to coordinate solutions to  cross-border systemic risk (see this paper I wrote with Zdenek Kudrna on what that meant for Eastern Europe). The implication: to curtail this mechanisms of interconnectedness a la IMF recommendation, regulators should seek to fragment global banks  into self-standing, domestically financed units.

2. Transnational banks – central bank policies.

As the IMF now recognizes, central banks in developing countries with open capital accounts have legitimate reasons to choose, and often do, a two instruments-two targets framework. Interest rate policy targets price stability while sterilized currency interventions prevent the exchange rate from appreciating excessively in the presence of large capital inflows. Sterilized interventions typically connect two markets: the foreign currency market where central banks first purchase (or sell) foreign reserves and pay by creating domestic money; and the interbank money market, where the central bank absorbs the newly injected liquidity in order to maintain its interest rate target (otherwise interest rates would fall).  For example, China sterilized 90% of its foreign currency purchases over 2000-2009 . 

Aside from the costs to the central bank, sterilized interventions offer transnational banks new modes of profit generation (carry trades), and perversely increase capital inflows. Transnational borrow abroad  at low interest rates and sell foreign currency to the central bank, to then place the domestic liquidity resulting from that sale in risk-free sterilization instruments. Downside risks are limited because these sterilization games appreciate the currency. Because banks ultimately decide how much liquidity they want to return to the central bank, sterilized currency interventions in fact provide transnational banks with ample liquidity to be placed in high-return domestic asset markets, with asset bubbles funded through growing cross-border exposures. The BIS warned that such perverse effects are likely unless the central banks bypass the banking sector all together in its sterilization operations.

3. Transnational banks - non-resident investors.

Non-resident actors that hold domestic assets typically do so as part of carry-trade strategies: cross-currency leveraged borrowing at low interest rates to place in high yielding assets. Carry-trade are notoriously volatile because non-resident investors are exposed to two types of risk: the risk that the target currency will depreciate, wiping out yield differentials and the risk that liquidity conditions in the funding currency become tighter (as in a crisis). Under such conditions, carry trades unwind rapidly, accelerating currency depreciation and bursting bubbles in the asset markets that they had targeted. Carry trades embed the most problematic features of globalized finance: difficult to measure, difficult to monitor, highly dependent on leverage, exacerbating cross-border exposures, speculative and pro-cyclical. 

To hold domestic assets, non-residents rely on domestic banks to provide funding in local currency. In derivative markets, resident banks take the other side of non-resident investors, earning profits from arbitraging the spot/forward markets. Regulatory authorities that sanction this relationship do so because non-resident investors improve the liquidity of domestic asset markets during boom times, and ease funding constraints for both private actors (including corporations) and governments (in the sovereign bond market). Those that dont because concerns for exchange rate volatility trump the drive for financial deepening, can impose residency-based capital controls (as many have done since the crisis) or follow the IMF position that 'restrictions on nonresidents‘ access to funding in local currency can at times make currency speculation more difficult’, as Singapore in the early 2000s. The key to breaking this type of fragile complementarity is to prevent local banks from providing domestic funding or engaging in interbank derivative trading with non-residents.

This is the graphic representation of  fragile institutional complementarities, the risks they generate and the types of policies regulators have already put on the table :

Putting ideas in practice 

Whereas IMF research on interconnectedness does engage with fragile complementarities, its policy advice retains the early portrayal of financial globalization powered by a handful of large global banks as an essentially benign process.

The Vienna Initiative 

In the case of the Vienna Initiative, the IMF agreed that foreign-owned banks should fund locally to avoid the possibility of future abrupt withdrawals triggered by problems of the parent bank. Yet it rejected suggestions for segmenting cross-border banking, instead proposing better ex-ante home-host coordination and market-based solutions to cross-border exposures: host regulators take the necessary measures for financial deepening that would improve the availability of domestic savings. It objected to discrimination in liquidity support (used by the Romanian central bank to defeat a speculative attack led by non-residents and funded by local banks) and to any crisis management measures that would increase the taxation burden for banks (used by Hungary after 2010). Its advice in the Vienna Initiative  did not consider the systemic risk underpinning the presence of non-resident investors, quite the contrary. Its recommendation for financial deepening implicitly sanctioned the presence of non-resident investors.

Transnational banks/ local banks
Transnational bank /domestic macro-policies
Transnational banks/non-resident investors
IMF research on interconnectedness
internal capital markets increase transmission of macro-financial risks btw home/host: fragmentation desirable
*transnational banks have complex business models; active asset and liabilities management;
*sterilized currency interventions necessary
curtail non-resident access to funding to prevent speculative attacks
Vienna Initiative
Ideas about financial linkages
*cross-border funding problematic
*domestic banks should fund locally
*investment opportunities for banks
*adverse tax/regulatory adjustments
Policy suggestions
*NO segmentation of cross-border banking (capital controls)
* ex-ante home-host coordination
*financial deepening
*no discrimination in liquidity support
*avoid fiscal costs for banks
Bilateral appraisals
Ideas about financial linkages
*cross-border banking essentially sound (Mexico)
* sustainable local banking model (Colombia)
*spillovers from European-owned banks (Chile)
*aggressive sterilized interventions costly for central bank (Mexico, Peru)
*transnational banks supportive of local capital markets
* high-dependence on non-resident capital entails systemic risks (Mexico, Peru)
* non-resident increasing dominance in stock market; effective restrictions on non-resident presence in sovereign debt market (Colombia)
* Chilean peso: funding currency for carry-trades into BRL
Policy suggestions
*costs from Basel III
and global regulation of transnational banks (Mexico)
*monitoring of cb balance sheet (Mexico)
* no actor-based analysis of central bank sterilization instruments
No analysis of Peru’s 2010 residency based capital controls; or Colombia’s reduction of withholding tax for non-resident portfolio investment in 2012

Such conventional IMF policy advice in the Vienna Initiative may  simply reflect the politics of international bureaucracies since these were EU negotiations involving the European Commissions, the ECB and the EBRD aside from home and host regulators. Transnational banks were to turn this complex regulatory terrain to their advantage and minimize the impact of those negotiations on their preferred business models. 

Bilateral advice in Latin America 

However, bilateral policy advice since the crisis paints a similar picture of little or no policy advice to curtail fragile complementarities in a selection of countries where policy makers expressed public concerns with the systemic risks these entailed. Policy documents recognize fragilities, discussing  possible contagion from the European crisis in countries where European (Spanish) banks have an important presence; recognizing the costs of massive sterilizations for central banks, and the systemic risks associated with the dominance of non-resident investors in sovereign and private debt markets. Yet policy suggestions oddly fail to draw the implications from these identified risks. In Mexico, the IMF warns that Basel III (and additional regulations for global systemically important banks) may have adverse consequences for domestic capital markets. For sterilizations, the IMF advises a careful monitoring of central bank balance sheets but without any concrete policy measures or indeed any actor-based analysis of sterilization policies. In what concerns non-resident investors, there is no attempt to examine capital control measures based on residency that both Peru and Colombia implemented since 2009.

In sum, what the analysis reveals is a better theory of systemic risk underpinning interconnectedness but light touch in policy practice. This is an interesting instance of an idea that becomes institutionalized but  does not shape policy practice. 

No comments:

Post a Comment