Leaping the Chinese wall

I don't know if its true, but I've heard it said the Great Wall of China was more efficatious in keeping the Chinese in than the Mongols out. And that seems also to be the thought behind the notion of "Chinese" walls within entities who otherwise share common information and interests.

I prefer the idea of a cone of silence, because it more clearly focuses on the silence concept and technological means to achieve it; walls just seem made to be destined to be breeched or jumped by determined humans. As happened with Citigroup's ineffective Chinese wall between its advising group and its proprietary trading group.

And that has been at the heart of some analysis in the Herald today, questioning whether there is an inherent conflict of interest between advising and trading, and if there is, whether such a conflict can be managed. As I pointed out before, the situation looked to me to more properly be analysed as a case of market manipulation or insider trading because the conflict notion would not have arisen had Toll's interests aligned with Citigroup's notwithstanding the trading instructions having come from the adivsing section.

But that idea hasn't occured or otherwise taken traction with the powers who print their analysis. Anyway, Elizabeth Knight seems to take the view that conflicts such as appear from the Citigroup matter can be sorted and the conflict provisions of the law can be made to work in this kind of case, but there are bound to be problems with reliance on Chinese walls :

THE Australian Securities and Investments Commission has a real problem in successfully prosecuting insider trading. But, thanks to legislation passed last year, it now has the legal armoury to force investment banks to manage conflicts of interest.

These new laws on conflicts of interest are what ASIC is using in its test case against Citigroup. But it could just be that the regulator is taking rather a circuitous route to deal with the bigger issue of insider trading.

It might sound like the mother of all conspiracy theories but, if you think about it, if Citigroup or any other financial conglomerate has a discrete division that buys and sells shares on its own account to make money and another that acts in the interest of a client, then it could have a conflict of interest.

Theoretically, there was no chance of insider trading. It was not until a manager at Citigroup noticed the firm's trade in Patrick shares and issued an instruction to stop buying Patrick shares that he opened up the firm to allegations of insider trading. This is ASIC's contention.

ASIC is alleging that the selling he undertook after he was told to stop buying was insider trading. And this will be a very tricky case to run.

But here is the point. If Citigroup had in place a proper mechanism for dealing with conflicts of interest then perhaps insider trading would not be such an issue.

Let's take for example a mechanism whereby Citigroup, or any other bank for that matter, instituted a policy whereby once it took on a corporate client it banned its proprietary desk from trading in the client's shares or the shares of any other company that was involved in the deal, which in this case would have been Toll's takeover target, Patrick. There are some investment bank that already have such a structure in place.

In this instance, Citigroup would still have been able to buy and sell on behalf of outside clients but not on its own behalf. Once the traders became aware of the ban they, of course, would be aware that something was going on but they would not have been able to profit on the investment bank's account.

And, given that proprietary traders are remunerated according to the amount they make for their own organisation, the incentive to trade on some level of inside knowledge would disappear.

(Of course, they could always tip their favourite clients into banned stocks but, having official knowledge that there was something happening would make this harder to justify.)

What ASIC and other regulators around the world are attempting to achieve is a transparent market. The creation of Chinese walls is not a particularly effective way to achieve this.

Because as time and share-trading activity have shown for a very long time, there are always holes in the walls. But as I have said earlier, making an insider trading case based on these perceived holes has been difficult. But if the regulator can mount a test case with Citigroup around conflict of interest that forces transparency industry wide then it closes off a large avenue to insider trading.

And the reality is that the real issue of insider trading is not about John Smith getting a wink and a nod from this brother-in-law and taking a $10,000 punt on some stock.

It's about big institutions like investment banks trading in the grey zone.

If the court decides that Citigroup does not have an effective means of dealing with conflicts of interests and makes a ruling that raises the bar for disclosure, then a big wad of suspect trading from any number of investment banks that occurs around a major price sensitive announcement might just disappear.

For this reason alone the ASIC experiment is worth a try.

Alan Kohler takes a look at the degree of proprietary trading in Australia and considers it "excessive", which is interesting because, as he describes it, it is nowhere near the extent of the big US houses, as described in the post Risque business. He also casts aspersions on the Chinese wall defense:

THEY all get up to it, but the idea that proprietary trading is an essential part of investment banking is nonsense: there are plenty of large integrated investment banks that don't do very much at all, writes Alan Kohler.

Macquarie Bank, for one, says its proprietary book is kept to the least necessary to sell warrants and options because, apparently, the bank believes principal-trading is not really compatible with advising clients and flogging things to them (for a small fee).

But unlike Morgan Stanley, Macquarie is keeping its head down in the stoush between the Australian Securities and Investments Commission and Citigroup over the latter's alleged conflicts of interest and insider trading, and won't comment.

Morgan Stanley's local CEO, Steve Harker, has declared that there is too much proprietary trading going on in Australia and that investment banks (other than his, presumably) are doing it to inflate their market shares. Others in this gossipy business are muttering snippily that Harker is just trying to do a bit of market share winning of his own.

Actually my quick and dirty survey of a few annual reports suggests that Citigroup - now the uncomfortable defender of proprietary trading by investment banks - is less of a trader on its own account than Morgan Stanley. Its average global "value at risk" (VaR), which is how you measure these things, was $103 million in 2005, or 0.4 per cent of its net profit (to get an idea of its scale within the bank, the percentage of profit is not especially relevant). Whether Citigroup Australia pins back its ears and runs a bigger book here is not known.

Morgan Stanley obviously doesn't do it in Australia on Harker's watch, but its group VaR last year was $93 million, 2.2 per cent of net profit. Macquarie Bank had an average VaR of $16.89 million, which is supposed to be about the least you can do while offering warrants and options, although it's still 2 per cent of Macquarie's net profit.

And the biggest principal trader? It seems to be UBS, although there's not a lot in it. Its balance sheet shows an average VaR in 2005 of 346 million Swiss francs ($371 million) - 2.5 per cent of net profit.

It's clear that only a small amount of proprietary trading is necessary to look after clients and you don't need to use it as a profit centre; the same goes for foreign exchange trading by banks (just ask NAB). In general there is little doubt that proprietary trading by all financial institutions that are meant to be there to serve clients is excessive.

If ASIC wins it will produce a revolution, especially if the judge finds that any proprietary trading at all is incompatible with the new conflict of interest laws (which is possible - judges can be unpredictable creatures). If that happened, the only way to prevent a stampede of investment bankers through the departure lounge at Mascot would be to ensure that regulators everywhere copied Australia, which is not entirely impossible.

ASIC has already been contacted by several regulators around the world interested in the case, although not by the Securities and Exchange Commission of the US. Given the ferocious attack in the US on conflicts of interest between dealers and analysts, it might only be a matter of time before that happens.

Meanwhile a settlement between ASIC and Citigroup looks a long way off, especially since ASIC has taken such a hard line and decided to bypass the fireside chat stage and go straight to court, which has made steam emanate from Citigroup ears and expletives echo around 399 Park Avenue, NY.

ASIC appears to be trying to create a debate (success there already) followed by a result that will bring clients more into the loop on proprietary trading and conflicts of interest. It seems that if Toll had had an opportunity to discuss the prospect of Citigroup trading in Patrick shares, and perhaps lay down some guidelines, this case would not have been brought at all.

It seems to be saying that Chinese walls are not enough: clients have to be given the chance to express a view about any conflict and perhaps set out some ground rules - including no trading at all in that case.

You may have noticed I post a lot of Kohler's stuff, so I consider him pretty cluey. But I hope he's not on the winning track with his apparent suggestion that ASIC wants client input on a case by case basis for determining what is and is not a conflict of interest. If Toll had waived the conflict in this case, for instance, wouldn't there still be an arguable case of insider trading or market manipulation, and wouldn't that still be difficult to run without the help of the conflicts legislation?