Governments are busy formulating regulations for the financial services sector to ensure that renewed exuberance does not result in systemic risk of the sort we have witnessed over the past two years. There is a tough balance that needs to be struck between implementing sufficiently tight regulation to curb excessive risk-taking and being so restrictive that innovation and efficiency are compromised.

One topic that is currently under examination is the appropriate size of firms in the financial services sector. How big is too big? Well, the market is supposed to sort out the matter of optimal size, except that firms appear to have grown larger than can be justified on the basis of efficiency gains from large-scale operations.

Other than the objective of limiting monopoly power, there is an even more significant public-policy issue involved. In a sector whose health is crucial for the smooth functioning of the entire economy it is important to minimise the possibility of systemic risk.

When an institution is "too large to fail" it poses a dilemma for the authorities. They can err on the side of caution by imposing a lot of restrictions which may also hobble innovation and efficiency. Alternatively, they can allow firms a good deal of leeway which, unfortunately, increases the risk of aggressive behaviour that may get them into trouble and necessitate being rescued at taxpayer expense.

One way of getting around this problem is to restrict the size of financial firms so that the failure of one or two does not pose systemic risks and can easily be handled by regulators. Naturally, interest groups will fight hard to maintain soft regulations and avoid size restrictions. Especially since authorities have demonstrated by their past actions that they are always willing to save big institutions from failure

Some firms, such as Citigroup, have grown enormously over the past decade becoming large conglomerates that are difficult to manage. The hoped-for synergies among different divisions were never fully realised in the process of empire building.

More importantly, from a risk management point of view, the top managers aren't always aware of the magnitude of their total exposure to risk on a global scale. The problem for such companies is that if they centralise decision making, this may result in bad decisions because those calling the shots are too far removed from local conditions to take effective actions.

However, if they decentralise authority, they may lose the ability to properly assess how local decisions add to the risk faced by the entire organisation. Why not strike a happy balance between centralisation and decentralisation, you may ask? A nice thought, except that it is difficult to achieve in practice.

In a large bank there are many individuals and groups engaged in risk management functions, located in various divisions and operating units. The problem is how to co-ordinate them, allowing timely analysis and information flows, and rapid decision making. In a small firm this task is achieved more adequately than in a large one.

What's more, the larger the firm, the more likely it is that organisational politics plays a major role in risk management. And, of course, the direction given by top management determines the extent to which risk issues are treated seriously or swept under the carpet.

We digress from our principal topic to address the issue of the crisis in Iran because of its important implications for stability and the regional balance of power. The battle lines have been drawn between reactionary clerics on one side and moderate clerics and progressive forces on the other. What is at stake is the survival of the basic institutions of the Islamic Republic.

In the wake of the revolution that overthrew the Shah thirty years ago, the clergy grabbed power ruthlessly under the leadership of Ayatollah Khomeini. There was no precedent for this in history or theology. The institutions that were created were entirely new.

Right from the start, the institutional structure was flawed; a theocracy combined with a measure of democracy, making for a difficult co-existence. But the concession to democracy was not an afterthought. It was a crucial in partially legitimising the system via an expression of popular will.

It has to be mentioned that the clergy have never constituted a monolithic body. There has always been a diversity of opinion among the Ayatollahs. Many of them were entirely opposed to the assumption of power after the fall of the Shah, preferring to leave government in the hands of secular authorities.

The current crisis reveals widespread opposition to the regime in all regions of the country and among all social groups. And the hardliners in power are responding by jettisoning the last vestiges of democracy and resorting to brute force. This essentially destroys any remaining legitimacy of the entire system set up by Khomeini after 1979.

The reactionaries have two choices: either make concessions now and allow reforms or blindly refuse any compromise and risk the eventual destruction of the Islamic Republic. It looks very much like they have made the second choice. Many senior clerics are well aware of the dire consequences but do not, yet, appear to have sufficient power to oppose the Khamenei clique.

For the US, the crisis presents both risk and opportunity. Risk, because instability in Iran may have spill-over effects in an already volatile region. Opportunity, because of the possibility of allying with progressive forces and tipping the balance of power in the region.

Currently, the guarded pronouncements made by the Obama administration are deliberate and entirely appropriate. Past US interventions in Iran have been on the side of reactionaries. Overt support for the reformists will damage their cause and help the hardliners. The Americans have to play it very carefully, and there is much at stake.

Iraj Pouyandeh is a strategist and senior portfolio manager at LOM Asset Management. He manages the LOM Global Equity Fund. For more information on LOM Asset Management please visit www.lomam.com