In today's market, trading has become complex and more automated than ever before. Not only is pre-trade activity and price generation data-driven and automated, so too is risk decomposition and basket transparency. Data accuracy can determine how to trade, what to trade and whether to trade at all. Exchange Traded Funds (ETFs) have grown rapidly in the last few years in terms of assets, volume, data accuracy is particularly significant for ETFs, which now have more than €650 billion under management in Europe alone.

In Europe there are around sixty ETF providers and nearly nine thousand individual ETF listings across thirteen markets. There is no conformity in data publication (remarkably, in some cases even at provider lever). Absorbing, standardising, validating and publishing this data - and in a timely manner - is therefore enormously important. 

As a trader it's your responsibility to correctly price, hedge and trade a potentially very large number of ETFs, which are typically complex products. Due to the sheer scale of your operation you will be depending on an automated work-flow where automated computer programs are importing and updating data that is beyond human faculty to manage. Your IT team has built processes that read data directly from ETF issuers, in-house analysts or third-party providers, after which they are collating this data and providing it to the front office. This presents a huge operational challenge, not so much because of technology, but because a minor error at the source can very quickly cause huge trading losses. How do you find a needle in a haystack?

Let's think for a minute about errors and their impact, in particular on the liquidity providers not forgetting of course that errors whether 'suspected or genuine' draw on trading desk resource time, a costly overhead.