The MEED survey also demonstrates that their playing field is rapidly expanding. Aggregate project-related and asset-backed financings exceeded $13,000 million in 2002, a substantial step up on the $8,100 million completed in 2001. With a further $4,700 million worth of credit extended to regional governments and financial institutions through loans and bonds, it is clear that the market has obtained critical mass.

The Middle East is coming of age as a structured finance market - it is no longer just a fascinating frontier market - but, in the process, the rules are changing and with them the key players.

There is a certain irony in the fact that as the region generates an ever-greater proportion of global project finance turnover - some estimate the region will account for between 20-25 per cent of total turnover - a number of heavyweight international banks are exiting the market and others are becoming more selective.

Of course, there is more than one explanation. Some of the market's original architects have undergone strategic repositioning. Citibank stands as a good example: last year it came second in the inaugural MEED league table of GCC lead arrangers; this year it fails to make the top 10. Judging by its deal flow and bidding tactics, it would appear that the decision has been taken at a senior level to focus on financial advisory work rather than lending into the region. In the short term, the tactic seems to be working. In recent months Citibank has continued to pick up interesting mandates - in Oman it is advising on the third liquefied natural gas (LNG) train and the proposed aluminium smelter project in Sohar.

Citibank is not alone: the likes of Deutsche Bank, ABN Amro, Credit Suisse First Boston and a fistful more have grown increasingly reluctant to use their balance sheet in the region, though all still actively pursue advisory mandates.

However, retreat on one front is matched by advances elsewhere. The regional banking community has been aggressively pursuing greater participation in the project and structured finance market. Driven by abundant liquidity - one of the reassuring side-effects of ongoing high oil prices - and a desire for the fees dispensed to those active at the senior levels, the regional banks have been booking their places early at the lead arranging table.

The result is a fundamental shift in the shape of the market. As the MEED survey shows, the international banks underwrote 39 per cent of the GCC's total deal volume in 2002. The statistics underline the retreat: in 2001 they underwrote 58 per cent, and in 2000 the figure stood at 64 per cent.

In the short term, the transformation of the market is undoubtedly a healthy development. The local banking community is learning new structuring and distribution skills and building capabilities that will be of profound use in the future. They are also diversifying revenue streams - the sought-after elixir of fee income is being consumed.

However, there could be trouble ahead. The two main concerns both centre on ongoing capacity. One stems from the limited ability of regional financial institutions to manage asset/liability tenor mismatches. Power deals in particular - with power purchase agreements that stretch out to 25 years - require very long tenor debt: some projects in Abu Dhabi have taken 20-year money, for example. But with no developed regional bond markets - only four banks in the region have issued paper and none of it has maturity beyond five years - the local banks have no access to equally long-tenor liabilities.

Attitudes towards such balance sheet imbalances vary. Some banks appear to have simply ignored the problem, but most seem to have adopted a strategy of lending up to their own shareholders' equity levels and beyond that taking a view on the average life of certain aspects of their deposit bases. So far, this manufactured approach has worked and no banks have run into serious balance sheet management problems. However, the tenor mismatch issue does impose some sort of a limit on the ability of the regional banking community to absorb ever-growing volumes of long-tenor project finance assets.

It is worth noting that the aggregate shareholders' equity of the top 15 regional lead arrangers stood at less than $20,000 billion as of the end of last year. The ceiling has not yet been reached, but it might prove to be a little closer than some expect.

There is also the danger of a more general liquidity squeeze. The high oil prices that have been sustained since the 1998 shock have brought revenues into the coffers of regional governments which have cascaded into the banks and been readily available for big ticket lending. However, bankers can suffer from collective amnesia and although memories of liquidity squeezes are distant, this is no guarantee against their recurrence. What can be forecast with greater ease is that if pressure is brought to bear, long-tenor, low-margin project lending will be the first activity to be discarded.

For now, such dangers seem some way off and, as the MEED survey demonstrates, the local banking community is now providing the base load of the region's underwriting for the first time. Unsurprisingly, this is having an impact on pricing patterns. It has long been the lament of international bankers that the regional players dictate the price while they do the structuring work. And there is probably some truth in it. For the international banks, too many deals have step-up margins that start on the wrong side of 100 basis points over Libor. Certainly, margins have remained stubbornly low and the competition from the local banks for still fairly limited deal flow has been one of the determinants. The borrowers - and the regional economies - are the beneficiaries, but the party will not last forever.

There is a strong likelihood that the crisis in Iraq could prove a turning point. While the project finance teams of international banks sitting in Bahrain and Dubai will almost certainly retain their faith, the possibility remains that credit committees back at headquarters could take a dimmer view of lending into what they might regard as an unstable region. The trend of international banks retreating from the region - given impetus by the aftermath of the 11 September atrocities - could accelerate further.

Significantly, not only could there be an evacuation of lead arrangers but also the disappearance of those international banks that have been content to enter deals during general syndication. Murmurings of a disappearing 'stuffee' market have increased over the last six months and the more it shrinks the greater the pressure on regional syndication. The point may well be reached this year in which the point of demand/supply equilibrium is reached and, if it is, there could be a sharp readjustment in the margins on deals. There is also the possibility that some of the underwriting fees - long regarded as the cream in many regional deals - will really have to be earned in the months ahead. It is difficult to predict exactly when pricing patterns will start trending upwards, or by how much, but the movement could be sudden and significant.

The perception of syndication risk which accompanies regional instability has grown increasingly real and has prompted the evolution of new deal shaping tactics. Financial advisers and lead arranging banks have been looking to mitigate this risk through the deployment of multiple tranches aimed at different liquidity pools. For example, the $1,550 million package for Aluminium Bahrain is made up of a combination of commercial debt, a metals trading package, a local currency bond, an Islamically-structured facility and possibly a loan from the Japan Bank for International Co-operation. A similar, if slightly less complex, approach is expected to be taken on the forthcoming Umm al-Nar independent water and power project in Abu Dhabi. Plans have been drawn up for an Islamic tranche and multiple commercial debt tranches with varying tenors.

Another tactic - pioneered in Qatar during the last liquidity squeeze - is to form large lead arranging groups as an effective means of mitigating syndication risk. The Oryx GTL deal stands as an extreme example: 15 lead arrangers were mandated for the $700 million deal and, as a result, when the deal came to selling down, all of the group started below its hold target - a definition of low-risk syndication. The model has been copied elsewhere, with Saudi Arabia's Jubail United Petrochemicals Company and the refinancing of Oman LNG following a similar route.

However, the tactic is not always used. Many bankers were surprised in February on receiving the preliminary information memorandum for the Sohar refinery debt package to read that a lead arranging group of only five to eight banks was envisaged.

There are other mechanisms that remain unexplored. Perhaps the most interesting is the development of a secondary market in structured assets. A liquid market would act as a backdoor exit route for regional and international banks from long tenor deals. And this possibility alone could ease their entrance through the front door. Ironically, one of the reasons why such a market has not developed has been the robustness of regional projects.

With a small handful of exceptions, the project finance initiatives have progressed according to plan, reducing the appetite for discounted debt sales. There has been some secondary market activity - mainly related to Saudi exposures - as some of the Japanese heavyweights went through a period of balance sheet restructuring following a string of mergers in Tokyo, but otherwise the market has been thin. However, there can be little doubt that the primary market would benefit if a secondary market were to mature.

Such developments probably remain some way off. For now the scene is dominated by concerns over regional insecurity. With no shortage of major deals queuing up to come to market, the next nine months are sure to hold excitement and sleepless nights for the region's financiers. It is at such times that deals become harder to do and the bankers really start earning their wages.