A Lesson from Abroad: Pre-trade Risk in Brazil

Thanks to some very competitive exchange prices for day trading, the high frequency trading (HFT) community has started to set up shop in Brazil. Current estimates suggest that 5% of trades on the stock exchange BM&FBOVESPA can be attributed to high frequency traders and, with a policy in place to attract more, that number is set to increase.

The jury is still out about the benefits that HFT brings to any market. Its proponents will argue that high frequency traders provide what amounts to a market-making activity that creates valuable liquidity, and has been credited with reducing bid/ask spreads and making markets more efficient for everyone.

On the other hand, HFT firms are not obliged to always make liquidity available and so have been criticised for pulling out of markets when conditions are volatile. The naysayers point to the dramatic fluctuations in price in the US, and lay the mini-crashes firmly at the feet of high frequency traders.

But it’s not clear whether this is a case of causation or correlation. The more likely culprit is not HFT per se, but naked access (which is now in the process of being banned in the US) and the more general failure in some quarters to manage pre-trade risk. It is here that Brazil, as a later adopter and long-time observer of the HFT phenomenon elsewhere, has a distinct advantage.

From the outset, the Comissão de Valores Mobiliários (CVM) has put robust pre-trade risk checking requirements in place, and has never allowed naked access. It has encouraged the exchange to put circuit breakers in place to limit the contagion of risky trading activities. And it has specified a number of risk checks that must be performed pre-trade, and applied different parameters according to the liquidity of the stocks concerned.

Direct Market Access

The more pertinent question in Brazil therefore is not what checks should be carried out, but where and by whom. The country has four distinct direct market access (DMA) categories. In the first category (DMA 1), buy-side clients use their broker’s existing connectivity to the exchange. In the second (DMA 2), which is more of a distribution model, buy-side firms use an authorised provider to access the exchange, which allows them to maintain a roster of competitive brokers. However, both of these models place the burden of responsibility for checking, configuring and controlling their clients’ limits with the broker. In both cases, the checks performed by the broker must mimic those conducted at the exchange.

In contrast, the third and fourth categories, where buy-side firms connect directly using their own infrastructure (DMA 3), or work on a fully co-located basis (DMA 4), the exchange provides the pre-trade risk control tool themselves. These are the areas that have traditionally been of concern, as the client order goes direct to the exchange without the broker intervening. HFT falls in the DMA 4 category. Increasingly, firms are opting for proximity hosting over co-location which is closer to a DMA 1 or DMA 2 scenario because it uses a broker data centre. Consequently the responsibility for pre-trade risk checks shifts back to the broker.

The question here is which method is preferable in this rapidly changing environment. The exchange will argue that by performing risk checks itself it creates a fair playing field since all players are working with the same level of latency. On the other hand, brokers point out that this is a valid source of competitive advantage and they need to have more control over their clients’ risk parameters.


The debate is ongoing, but what seems clear is that there is also a degree of future-proofing to be gained from broker-level risk checking. Certainly it enables brokers to apply broader risk checks on top of those mandated by the CVM, and to take control of their own risk parameters both at a client and individual order level. What’s more, it enables brokers to co-mingle equity and derivative risk checks at a firm level, which is a significant advantage.

Most importantly, brokers can conduct risk checks over multiple venues, which is an essential feature in a fragmented market. With BATS openly discussing a potential launch in Brazil, and several other consortiums in play, this will be the critical issue, not just for the HFT community but for all other players in the Brazilian market. The potential for changing the risk equation in this marketplace is enormous and the debate around the best solution will continue to rage.


Related reading