The Gateway to Algorithmic and Automated Trading

Black box in the Bloomberg box

Published in Automated Trader Magazine Issue 13 Q2 2009

Amid all the talk of ultra high frequency trading and black boxes, automation tools for human traders are often overlooked. Yet such tools can hugely enhance productivity and profitability. Gary Stone, Director of Trading Research and Strategy and Brian Coffaro, Manager of Derivatives Development at Bloomberg Tradebook, explain how and why.

By the very nature of their job, professional traders spend their lives looking for profitable anomalies in market behaviour. These can be anything from the way a particular market reacts to a certain type of news event, to a leading/lagging correlation between two markets where a move in one market is followed by a commensurate move in another.

One way of exploiting these often short-term anomalies is to attempt to code them as automated trading models, but an obstacle to this approach is the time required for model development and testing. If the trader who detects the anomaly is not also a programmer, then he/she will have to hand off the development process to someone who is. Apart from the immediate opportunity cost, there is also the risk that by the time any resulting model is completed, the original anomaly may have disappeared.

A further complication is that the original idea may not be reproducible in the form of a quantitative model. Skilled traders are often processing a large number of inputs, many of them subliminally. Incorporating all these into an automated model may be impossible - even if the trader thinks to mention them all to the programmer in the first place. Such inputs are often interrelated in such a complex way that they are effectively qualitative not quantitative. For example, at what percentage above expectations does a particular economic number represent a buy? On one day, market conditions and 'feel' might indicate one value to a human trader - on another, a completely different value.

Leveraging the opportunity

From talking to many Tradebook users and other traders about these issues, it became obvious that an alternative 'semi-automated' approach was needed. Instead of black box automation, flexible tools were required that would enable human traders to set parameters for position entry/exit that could easily incorporate multiple inputs from multiple markets and data sources. These would enable traders to quickly set entry/exit parameters in accordance with their current market view, rather than having to be hard coded in a black box model.

At the same time, the workflow benefits of these tools would enable traders to scale up their activity across multiple opportunities and markets. Rather than being constrained by the limitations of manual trade management and therefore having to ignore potentially profitable anomalies, traders could use these tools to increase their trading capacity and/or their level of diversification - but without increasing their workload or introducing the possibility of errors.

In response to these needs, Bloomberg Tradebook now provides the answer in the form of its new Enhanced Contingent Orders and Event Trading Algorithms. Available across multiple markets (options, futures, foreign exchange and US or global
equities), these tools provide all the necessary functionality for traders to immediately and fully exploit all the trade opportunities they detect, not just a subset.

Crude + Euro = Profit (if you had the tools)

The European Central Bank regards energy costs as a major inflationary driver, so any increase in oil prices has been likely to trigger a rate hike, thereby strengthening EUR. Furthermore, with the Eurozone being a major oil importer and oil being USD-priced, strengthening EUR offsets higher oil prices.

This situation has resulted in a strong correlation between EUR and crude prices. A classic opportunity to exploit this relationship arose in July '08 when a battery of technical indicators forecast that oil was approaching a peak. By implication, any collapse in oil prices would see a commensurate drop in the Euro.

Figure 1
Figure 1

Tradebook Enhanced Contingent Orders are the perfect way to take advantage of this trading opportunity, as illustrated in Figures 1 and 2. The trader establishes an order based on two trigger factors:
• A break below the 142.49 level in the NYMEX crude future
• A divergence of more than 1m barrels between the consensus forecast and actual values of the DOE US Crude Oil Inventories report

If these conditions are satisfied, a short limit order to sell 5mn EURUSD is initiated. Once the ticket is completed, all the trader has to do is wait; the Tradebook inter market contingent order does the rest. Figure 1 reveals that inventory increases by nearly 3m barrels, versus a forecast decline of 2.2m. Price reacts accordingly and Figure 2 illustrates the profitable outcome in EURUSD...

Figure 2
Figure 2

Enhanced Contingent Orders

Enhanced Contingent Orders take the concept of one-cancels-other (OCO) orders to a new level. In addition to standard OCO, these include the following order types:
• Target - triggers an order at a specified price. Can be based on greater, equal or less than last trade, bid or offer.

• If Done - allows two orders to be entered but the second order can only trade after the first one is filled. Useful for scalping or other small profit target type trades. Also an efficient way of handling futures calendar spreads, where the less liquid contract month order must be filled before the more liquid one trades - thus minimising legging risk.

• IOCO - combines If Done and OCO orders so that the OCO order is conditional on the first leg of the If Done order. An efficient way of entering profit target and stop loss exit orders after an initial order opens a position.

• ECO Data - an order based upon a trader-specified value for an economic data release. The order ticket for an ECO order automatically displays the consensus forecast and previous value for the economic data in question. The fastest and most effective way for a human trader to trade off economic numbers.

A key feature of Enhanced Contingent Orders is that their trigger functionality is not restricted to just the market being traded; triggers can be based upon conditions placed upon a combination of up to three other markets or pieces of economic data. So, for example, a trade in gold futures could be initiated based upon triggers established on gold production data, EURUSD and an index of gold producers' stocks.

Tradebook's custom index creation facility also means that traders can base triggers and actual orders (via Tradebook's basket trading tool) on synthetic indices of their own devising, so there is effectively no practical limit to the sophistication of triggers and orders they can build. (For example, a leading/lagging correlation trade between a stock index future and a custom index containing some of its constituents.)

Event Trading Algorithms

Tradebook's Event Trading Algorithms build on Enhanced Contingent Orders by adding the concept of an Ultimate Limit. Particularly during high volatility periods (such as immediately after a major economic release) slippage on market orders can be substantial. To avoid this, many traders will typically use a Tradebook limit order pegged to the bid or ask. However, this still leaves the risk that the market may gap through that limit and the trade will be missed.
Traders don't want to be hostages to fortune by using market orders, but nor do they want to miss out on the trade completely.

Ultimate Limit orders are intended for just this eventuality; when completing the order ticket, the trader specifies two limit orders - one conventional and one Ultimate. In the case of a buy, the conventional limit will be below the Ultimate Limit, vice versa for a sell. If the market gaps through the conventional limit, then the Ultimate Limit becomes active; the ultimate ceiling/floor on the trade avoids slippage while simultaneously increasing the likelihood of participation.

Enhanced Contingent Orders and Event Trading Algorithms can be used on a standalone basis for small orders in liquid markets. However, for more demanding order execution conditions they can also be linked to Tradebook execution algorithms, such as Bsmart, Reserve Scaleback andTradebook VWAP/TWAP. So if all the trigger conditions are met, rather than simply firing an order into the market, the trade is handed off to the appropriate algorithm to ensure the most efficient execution.

Productivity, scalability, profitability

Particularly in the current market conditions, anything that enables traders to make bigger and better use of their skill sets without requiring a major investment is an obvious winner. Tradebook's Enhanced Contingent Orders and Event Trading Algorithms deliver on this promise by maximising the return on human traders' intellectual capital.

Both tool sets enable traders to quickly set up and control multiple trades, so trading opportunities that previously had to be discarded because of trade management capacity limitations can now be taken.

In addition to increasing potential returns, these new abilities can also be exploited to increase diversification and improve risk management. Furthermore, as increasing competition continues to reduce the shelf life of trading opportunities, Enhanced Contingent Orders and Event Trading Algorithms also deliver faster time to market to improve alpha capture.