Adapting your trading style – How the changing market landscape is driving new skills

Share

Appetite for risk has never been lower and liquidity has never been tougher to identify. The downstream effect is impacting every part of the electronic trading business and culture. Quod Financial’s Ali Pichvai examines where he sees the sea change in trading skills and style.
We are at the midst of a structural, and subsequently cultural, change in the capital markets. Firms’ appetite for risk is shifting and counterparty risk is now high on the agenda. This trickles to each function and aspect of investment and execution; from investment decision making, to risk management, and the mechanics of the electronic trading. In addition, liquidity is increasingly fragmented across a multitude of pools and is affecting how electronic markets are evolving. So how does this landscape impact how firms will trade?
Changes on the buy-side and how the future will look are still uncertain. The hedge fund industry, the great innovator investor class, has in large part been discredited and its model will need to drastically change. The quasi-demise of this large segment will leave a void that needs to be filled. It seems that the future lies in more transparent, better risk-managed, low-cost listed products, which respond to the appetite of global multi-asset investment and execution strategy of the investors. Furthermore it is now clear that liquidity and solvency are intimately linked, and evaporating or volatile liquidity creates systemic risk on solvency. This will, without doubt, have a large impact on future capital market structures.
The buy-side transformation will inevitably accelerate the pace of the current secular trends of more electronic trading on centrally cleared liquidity venues and competing global or regional multi-asset liquidity venues. NYSE Euronext, as a global multi-asset liquidity venue, seems to be the role model for all other market participants. The liquidity fragmentation, as observed today, will certainly be greater and more complex going forward. It also seems we have entered a second age of liquidity fragmentation, with three phenomena which have appeared, or been reinforced, in the current turmoil.
Liquidity is becoming ever more dynamic. As competition increases price wars are becoming more frequent, and pricing models are being altered to attract more and more liquidity. For instance, the rebate model for passive orders (i.e. by resting a passive order, you can receive a fee) has often been used as an effective marketing tool for new alternative trading systems. Clients are therefore moving their execution on a real-time basis from venue to venue, as pricing evolves within a competitive landscape, making liquidity ever more dynamic.
Liquidity is decreasing transparency. As new dark pools and brokers internalisation profligate, with the US equities having achieved 17% of execution in these dark venues, the level of transparency is decreasing. This creates a massive trading challenge. Transparent liquidity is important since it creates an efficient price discovery model, which then disappears into a non-transparent execution model.As transparency decreases, in addition to market data sourced from the different displayed prices, there is a need to move to real-time post-trade analysis, to rebuild a more intelligent picture of liquidity.
Volatility increases fragmentation and increases execution risk. The current intraday volatility, and an even lower period volatility, is much greater than at any other time, and bigger than the impact of incurred costs. As seeking liquidity becomes more important, fragmentation will increase. The result is a new type of risk which needs to be mitigated; the execution risk. This means that the investment case can be fully redundant if the execution in a highly volatile market is not properly performed. This risk evolves from the inability to execute down to execution too far away from the investment decision. Another obvious effect is that the widespread algorithmic trading engines, which were built to limit for low volatility markets, have become obsolete. Nowadays, in an averagely volatile day, it is not uncommon to have 300 basis points of volatility, which dwarfs a single digit basis point cost impact. That means that the next cycle of investment in algorithmic trading needs to be redirected towards liquidity seeking algorithmic trading (also called smart order routing – arguably a misnomer, since it is simply routing rather than delivering a real-time decision making process).
It is therefore reasonable to hope that the shift will mean that, at last, fully supported multi-asset trading becomes a reality. The change is more than overdue. The vast majority of current investment strategies are already multi-asset or even cross-asset, but are not properly supported by the trading and execution vendor, broker-based or even exchange-class systems. The pressure to respond to this unfulfilled investor demand, to need to holistically manage risk (for any given client across different asset classes, without being too lax or too restrictive), and to reduce the cost of overall trading systems and infrastructure, are potent drivers to make this conscious shift happen.
The mainstream sentiment, driven by the backlash to the financial meltdown, is that capital markets and trading are beyond control and need to revert to a simpler model. This is a simplistic response. In reality while the level of financial innovation increased, technology investment did not keep pace, and this was particularly true post the mild recession of 2001. It is unrealistic to have expected increasingly complex concepts, such as liquidity fragmentation, risk associated with execution and multi-asset / cross-asset trading to be handled competently with the same old technology. If history is any guide, the future will likely be even more complex. So those investing intelligently in technology will be tomorrow’s winners.
Looking specifically at the execution world, there is a dire need to apply new technology which can manage the current degree of complexity and the need to repatriate execution making within the firm, instead of fully outsourcing it to liquidity venues or the sell-side players. The good news is that adaptive trading technologies are becoming both available and affordable to most market participants.
Adaptive trading technologies deliver real-time decision making mechanisms which rely on real-time market data and post-trade data. They direct and seek liquidity whilst fulfilling the original execution objective (or best execution policy). This allows it to handle complex scenarios, with multiple execution policies and numerous liquidity venues, on a multi-asset basis. Realistically, these technologies need to leverage the current trading investment in terms of infrastructure, upstream and downstream systems, such as risk management, or connectivity, and be as non-disruptive as possible.
There are, however, potentially major failure points. These include:

  • Speed against complexity: There is a trade-off between the complexity of the decision making process (also called algorithm), and the effectiveness and speed of the algorithm. The right balance is usually empirically found on a case-by-case basis.
  • Testing the machine requires a new approach: The classical quality assurance and testing approach does not apply to these very complex decision making processes. The testing combinatory permutations for just three to four liquidity venues and some real life client trading behaviour exceeds 100,000s of cases. The cost and length of such extensive testing is both excessive and impossible. The only way is therefore is to take a new approach which is based on statistical sampling of the test cases (and the art of sampling is itself difficult) to simulate as closely as possible a real production environment.

One key application of adaptive trading technologies is the ability to address decreasing transparency. That means real-time post-trade analysis of non-transparent liquidity. This information can be then complemented by market data (assuming that dark pools are usually trading at mid-price) and used as a basis on whether to directly execute on a given dark pool or not.
In conclusion, the changing market landscape demands a shift to adaptive trading technologies and that shift needs to happen now. Firms are still investing in technology, albeit strategically, and in order to keep ahead of their competitors in tough market conditions, firms need intelligent trading systems that can adapt to a constantly shifting environment.