Exploiting Order Flow for the Discretionary Quant

Exploiting, Order, Flow, Discretionary, Quant, Trading Systems, fx trader, forex

The Discretionary Quant

How can we trade a market that is efficient enough to rankle most participants, yet inefficient enough to allow others to deliver huge profit?

To paraphrase Lasse H. Pederson, the market is neither perfectly efficient nor completely inefficient: the market is ‘Efficiently Inefficient’. Order flow is inefficient enough to create opportunities to trade a strategy for profit after cost, but efficient enough to discourage additional investing.

It is often said that the simplest strategies are the best. But how do you prove this concept? And how do we quantify risk?

Understanding market microstructure – order flow inefficiencies and supply/demand imbalances – presents opportunities to develop strategies which are not always math derivatives. It is possible to quantify a discretionary approach to trading price action without sacrificing either the opportunities presented by volatility, or the integrity of the statistics, by using a proof-of-concept approach as a basis for further development which is simple enough for both developing discretionary traders and automated models.

This article outlines the sufficient detail to support this approach, and is a practical outline in two parts.

Part One outlines discretionary price action trading using order flow inefficiencies and market structure behaviours. An agile, proof-of-concept approach is explained in order to test a basic, workable, automation-ready strategy for intraday trading to exploit these behaviours, including a straightforward, quantitative approach to modelling the strategy.

Part Two explores in more detail the expectancy & excursion analysis to deliver a ‘minimum viable strategy’ to trade, and includes an automation roadmap to accelerate specialisation and testing which can be adapted for a number of currency pairs. In the presence of the Cloud of Big Data, the discretionary trader can feel like the face at the window, with as much personality as a paper cup. So the article presents the pros’ and cons of automating the strategy for both testing and live trading.

Developing traders can use the test model contained here to answer specific questions that almost everyone asks: not just ‘where do I get in?’, but once in, ‘where should I get out?’?

Some commentators state that the edge in FX intraday trading is diminished by frictions (spread & intraday randomness) and is by default a negative sum game for most players. That’s why it’s important to manage the variance from concept: (a) understand market microstructure presented as price behaviours, (b) develop and test a strategy to exploit the behaviour as a proof of concept, and (c) meet and beat the concept expectancy as a discretionary trader or via a blend of automation and discretion.

Order Flow Inefficiencies

Price movement is a function of liquidity.  An inefficient price move is not necessarily a sign of volume; an absence of bids or offers can ‘spread’ price as it reaches for available liquidity, and this in turn can introduce additional order flow, creating a liquidity void in its wake. On exhaustion of these moves, price can ‘rebalance’ through this void to the last point of liquidity, a move that can be pronounced as institutional algos manage both inventory and adverse selection risk from the previous price run.

More specifically, market structure – higher and lower highs and lows – can act as a trigger for inefficient price runs, as price catches either a liquidity void or concentrated stops at these levels. This happens intraday as well as day after day (former Olympian and World Cup Bobsledding champion Chris Lori is by far the best practical resource on order flow inefficiencies in FX).