There's a lot of excitement, or perhaps hype, about China's capital markets. A French bank recently set up a joint venture in China, which will enable it to underwrite IPOs and bond issuance, and perhaps carry out brokering further down the road. SG has had a long presence in China: What's your take on how to tackle this market?

Poirier: There's always a temptation in business to relive past business models. In fact, investment banking is changing - it's becoming more commoditized. You can't relive the glory days of the past, however tempting. That's true for the industry globally and is reflected in the 100,000 jobs the industry has shed in the past couple of years.

Who is getting fired? They come primarily from among the salesmen and the analysts. Fund managers these days no longer listen to these once highly paid professionals. These days, funds follow the indices of a country and sector and trim or expand their holding strictly according to the amounts suggested by the MSCI or the Financial Times family of indices. A lot of these trading progammes are automated and computerized: there's no need for the input of an investment bank or brokerage. The fund company does it all in-house.

The old model is not dead yet, but it's changing. But a lot of plain vanilla products are automated. That's why Merrill Lynch and Goldman Sachs have now diversified their revenue sources.

How attractive is China within that paradigm? If you look at the figures, the country has seen 67 IPOs in the past year. In March alone, six companies IPO'd and Rmb 3.4 billion was raised. How attractive is that?

I'm not sure it's possible to make a profit from underwriting initial public offering in China, or from brokerage. There have been IPOs in the past year, for A and B shares, but the offer size is usually small, around $50. The March figure, Rmb 3.4 billion was down from Rmb 17 billion this time last year. Another limiting factor is that every securities house has a limit to the number of candidates it can submit to the China Securities and Regulatory Commission for approval. The most favored houses can only put ten companies forward per year. Fees are around 2-3%. How much money can you make consistently from that, given the fact there's over 110 firms competing for business - the majority of which are loss making, by the way, and driving brokerage fees down to zero? They are also state-owned by provincial governments, which means they can sit there and bleed for years before being merged or closed down. To be honest, 33% of the profit a loss-making security company is not very interesting to us.

Do you think the Qualified Foreign Institutional Investor scheme will bring about a major change in the perception of China?

I think it's going to take a while! So far you only have some banks applying for custody roles: How much real money is queuing up to enter the Chinese market? Remember, you can't take your money out for three years. On top of that, Chinese markets are volatile and uncertain, and the quality of companies listed is pretty poor. I don't see that as being very attractive to international investors, at least not yet.

There is a lot of change going on in the composition of the market. Last year, Zhou Xiaochuan, the then chairman of the CSRC, said that thanks to back door buyouts of the so-called non-tradable shares, the percentage of de facto private companies is rising all the time. Does that make it more attractive?

But I maintain the investment banking model is undergoing changes, and that's not going to change because of the presence of decent private companies. The securities business can't make it's way with plain vanilla products anymore. What do you bring to the table in what is increasingly not a specially complicated business. Maybe distribution.

Why?

The market is now very competitive. There is no margin. Take Research: it's very expensive. But everybody gets it for free, basically: too many people are competing for investor attention. But the buyers, the mega US investors, use only one or two brokers, often in-house. The buyers are shrinking the number of brokers they are using. Basically, there are too many brokerages in US, Europe, Korea and also in China. In Korea you can trade on the Internet for two cents per share.

So what is the model?

I don't need to own a stake in a company. I make a product: someone sells it. Why should I lock up money in somebody else's company? We provide them a good product, they distribute it, we both make money.

What kind of products?

I mean investment products on the asset management side. I want agreements, distribution rights. I can tailor products for my Chinese partner who sells the product, and we get a cut. I don't need a formal shareholding. I can use my capital for other things. I'd rather be on the side of the money collection process by selling products to the Chinese, than paying money out on equity stakes.

We are already selling guaranteed capital products denominated in US dollars to Chinese financial institutions, and that's already proving very popular. In addition, we have positioned ourselves with a Chinese partner - Baosteel - to set up an asset management business. It's not an outright acquisition, since we only have 34%, but it's a very convenient platform from which to distribute our product. You might call it an 'out of the money call option…!'

Last year you were also involved in project finance, one of them being for Vivendi. What does that say about your strategy?

Yes. The $250 million Vivendi acquisition of a stake in a Shanghai water company and arranging financing for the $2.7 billion Shanghai ethylene naphta cracker complex. Thats because we look at the Chinese market and we say: where is the money coming from? It's coming from Chinese need for FDI. So we position ourselves to advise companies carrying out joint ventures or acquisitions in China. We organize the financing. We don't tie up our capital. And that's a lucrative niche.