The New Risk Manager 2.0

24 February 2017
The Market Owl

For the last few months all of our efforts have been centralised on the launch of “DARWINEX reloaded”. One of the principal new releases of this launch will be the significant change in the functioning of our Risk Manager.

In this post we will explain the four principal releases in the new Risk Manager 2.0

  1.   Reduction of the VAR objective of DARWIN to 10%
  2.   Change in the formula to calculate the VaR of a DARWIN
  3.   New criteria of the closing of a position by the Risk Manager
  4.   Maximum D-Leverage by default for all the DARWINS

Before explaining the changes that we are going to introduce, it is important to understand why the RiskManager exists, and why we fixed the initial DARWIN risk at 20% of VaR.

Darwinex is FCA (UK) regulated asset manager, and therefore, has the ultimate responsibility of all the strategies that are quoted on our platform. As every underlying strategy of the DARWINS operates with an unknown risk, (and therefore out of our control ), if we are to permit that the investors replicate directly the underlying strategies of the traders, we would not be able to comply with our legal obligations as a risk manager for the investors.

For this reason, the need was there to devise a mechanism so that Darwinex can control the risk: the Risk Manager

So, how does it work? Very summarised: each time that the trader send an order to the market, his DARWIN ( through the Risk Manager) will replicate the said order in the investor´s accounts, but with a leverage controlled by our algorithms (adjusted as a function of market conditions at that moment). To reiterate, this is a VERY summarised version of how it functions.

Snip20170224_78

 

One of the principal advantages of the Risk Manager form the investor´s point of view , is that all the DARWINS operate with the same risk, allowing the comparison of every DARWIN on a like-for-like basis, with the need to incorporate normalisation ratios of their profitabilities.

On the other hand, with a defined risk, the Risk Manager assumes that the behaviour of the market will be similar to the other, more well known markets to the investor, such as the equity market. Therefore, if the Risk manager did not exist, a generic asset would not exist, ( if the trader operates with whatever risk he chooses, we would be in a similar situation to a market bazaar in which each stall holder would offer a distinct product , that would have no relation to anything else in the market: one would sell sweets, the next carpets etc)

So why a 20% VaR?

Once we have created the Risk Manager, we had to define the level of risk of the DARWINS. The search for what we consider to be an optimum level of risk for the DARWINS  has been ongoing since we launched our platform.

Initially, we decided to take advantage of the Beta phase testing of the platform, to offer the investors three levels of distinct risk ( 5%,10% and 20% of VaR). This would help us as an experiment to study the conduct of the initial investors, and from that base, to define the most appropriate risk level for our platform.

Snip20170224_79

After a test period of eight months, the main conclusions that we arrived at for the investors were.

They diversified their portfolios with an average of 6 DARWINS per account

They preferred the DARWINS with 10% and 20% more than those with 5%

We therefore discarded the DARWINS with 5%, but we continued to doubt which would be the best outcome , 10% or 20% VaR. Finally , we plumped for the 20% because with a diversification of 6 DARWINS and a 10% VaR per DARWIN, the risk of the account would round out at 2.4%, a value that would be considered excessively low compared to, for example, the median risk of the S&P (close to 4%)

Why therefore have we decided now to lower it to 10% with Risk Manager 2.0?

To change is to learn. During the last year , we have studied closely the behavioral changes of our users, and we have learnt that:

  1.   The DARWIN providers have ( or can have) loss aversion
  2.   An elevated risk in the DARWINS damages the positioning and the credibility of our market
  3.  A higher risk of a DARWIN produces poor timing for the investors
  4.  We can avoid that diversification reduces the risk of a portfolio excessively

Loss Aversion

During the last year, there have been cases in which the volatility of the DARWIN has generated changes in the trading models that the traders operate in their underlying strategies. Above all, the most conservative traders when it comes to risk management, have been affected psychologically by the higher DARWIN volatility, and have unconsciously modified their models, prejudicing the quality of the DARWIN, and consequentially their investors too.

This blog post on loss aversion is old but gold!

Market credibility

An elevated risk is not welcomed by qualified investors. Despite being completely distinct business models, on many occasions we have been compared to social trading platforms , which in most cases profit from their client´s losses (they have business models similar to those of the casinos)

Unfortunately, with a VAR of 20%, drawdowns of more than 30% are common, which can easily generate scepticism with regards to our ultimate goal of our users winning.

Paradoxically, the average investor prefers to make 50% with a drawdown of 15% than to win 100% with a drawdown of 30%. This is something that can be seen in the human psyche , but is not financially logical: from the strictly financial point of view , both investments are the same (in truth, the second option is better, as it generates a better return with less capital). Reducing the VAR to 10% a month , is , amongst other things, a bet to admit investors qualified in our markets, and thus will favour our traders.

Investors’ risk and timing

In the same sense, and to our obvious disappointment, we have learnt that many investors commit the same errors that almost all those active in finance commit: they tend to buy when the DARWINS are at their highs, and to sell at the moment that they produce drawdowns, producing systematic losses ( which is aggravating as the DARWINS with good fundamentals tend to gain in the long term and to recover quickly after they fall.)

This phenomenon merits it´s very own post , in which we will compare data and recommended models from the moment of their investment in DARWINS. In whatever case, reducing the VAR by a half, we hope to minimise this “panic effect” in our investors, making it more bearable to maintain their investments during the falls in prices (inevitable in whichever financial instrument is chosen)

Diversification, profitability and risk

Diversification is, without doubt,  recommended. To protect against the most unpredictable phenomena , such as the the “flash crash” in sterling on the 7th of October last year, diversification for protection against abnormal volatility spikes that our risk manager cannot prevent is vital.

That said, it is also certain that an excessive diversification results in less potential profitability. Accounts with more than 10 Darwins and a VAR of 10%, produce very reduced yields (the DARWIN investor does not characterise themselves as being very conservative). As the number of DARWINS increases , in our community we are experimenting with a growth in the size of the accounts, (and the quality of the DARWINS are also getting better). Because of this , we have devised a mechanism that will permit the investors to diversify without the need to relinquish profitability, but that is the subject matter of another post that we will write in the near future…It’s a surprise!!!

Lastly, we have no doubt that there will be users that will not be happy with the changes in risk, because they are happy with 20%. You cannot please all the people all the time. Or not?? To avoid dissatisfaction in Darwinex reloaded , we will introduce the possibility to use leverage on behalf of the investors , so that they can duplicate the leverage (and, in essence they can carry on investing at 20% of VAR).

Looking to the future, even though we are still in the conceptual phase, it is possible that we raise the available leverage factor as a function of the number of DARWINS in the account and their correlation ( although never more than a limit that we will settle on).

 

We leave it here , just the first few brush strokes of the new Risk Manager, and in following posts, we will explain in a more detailed form the other 3 changes mentioned in the introduction. As always we are always available for whatever you need on info@darwinex.com.