Big Data Technology for Investor Protection

Big Brother lives in Lausanne

Banque et Finance: Tuesday, June 11th 2013

Anyone who is interested in finance knows that although figures have intrinsic value, they can be made to say almost anything once interpreted by humans. It was following this observation that cfinancials, a young dot com company in Lausanne, decided to launch a product to measure, objectively and transparently, the safety margin of financial products. This measurement is based on the tracking and analysis, in real time, of a database of some 12 million financial products. The initiative was sufficiently innovative to arouse the interest of the European Commission which, since the 2008 crisis, has been looking for a monitoring system that would enable it to anticipate systemic risk. Its leitmotiv: to enable the EU to avoid plunging back into a situation even worse than the one it is currently struggling with. In fact, it has given this responsibility over to an independent structure, the ESMA (European Securities and Markets Authority), based in Paris. Along with the EBA (European Banking Authority) and the EIOPA (European Insurance and Occupational Pensions Authority), the ESMA has been put in touch with the Lausanne company.

Tiper (which stands for True and Intelligent Product Evaluator of Risk) aims, therefore, to “measure risk” based on purely mathematical algorithms which use the immense database of cfinancials, and not, as is very often the case, to “interpret risk”. There are as many interpretations as there are people on the planet, points out Michael Heijmeijer, the founding director of the companyThey vary enormously and are mostly subjective, based on the beliefs, on the degree of optimism, or even the motivations of their proponents. And yet, it is these interpretations that make the markets go up or down, and that play an essential role in risk-related decision-making.

Often, the success or failure of an investment is attributed to capital market trends and to the complexity of financial products. In reality, the markets are nothing more than a reflection of the appetite of investors. The products are, by their very nature, totally neutral vehicles. They give the results we want them to give. Michael Heijmeijer adds with humour: “f you spill your coffee, it isn’t the cup’s fault!. In fact, it is all about how these products are marketed and sold, with all the subjectivity this can generate. The current economic model is mostly based on the amount of the transaction and of the commission, and the issuers are never remunerated according to the product’s performance. Tiper cannot stop someone from making an irrational decision which, in fine, increases the risk of loss. But it can provide a reliable, objective and transparent measure of the safety margin they might take.

The way it works is simple: no more time spent analysing in depth the often complex structure of such and such a product, or the opinions of such and such an advisor or broker. The investor is equipped with a single index from 0 to 100. Financial products become ‘raw materials’ whose data and prices are collected and compared to calculate the risk level.
Real-time evaluation is made possible by a high-performance technology infrastructure and extremely efficient algorithms. Datamining, artificial intelligence and cloud computing make it highly reactive, on a par with the stakes involved. Investors and analysts should therefore find in Tiper a substantial helping hand so that the transgressions of 2008 and their social and economic consequences, which we are still feeling the effects of, do not happen again, and that so-called “toxic” products are, at last, identified.

Tiper in detail:

Tiper measures risk in real time and quantifies it from 0 (low) to 100 (high). To provide this evaluation, it analyses the following data:

  1. Absolute Risk Value • Distance to Total Loss (zero) • Standard Deviation based on Moving Averages
  2. Volatility • Fluctuations relative to Price Levels
  3. Liquidity • Daily average of cumulative number of transactions (CNT). • Comparison with the trailing 30 days moving against CNT. • Real Time Derivative of CNT
  4. Relative Risk • Peers. • Index
  5. Contextual Risk • Following risk mutations on ie. Structured Products

The IP is in the algorithms and artificial intelligence used to evaluate the Tiper. The technology is based on advanced datamining algorithms on huge databases.