We can notice that financial markets are subjected to the collective influences, to the psychology of the agents, to the diversity of the information following the role and the interests of each of the economic agents. This information constitutes partially the low(weak) signals of the market the impact of which on the Mark To Market varies according to other information which can be macroeconomic, stock-exchange and others. The fractal methodology uses a very few input data (date and Mark To Market) to identify the dominant low signals at the time of the calculation and to obtain relevant indicators which express the right price of the market. The methodology allows to understand better the breaks and the unexpected reversals. .
The calculated indicators are those commonly used in financial management. But their calculation is refined by the fractal methodology, which gives them a realistic sense.
The calculated indicators are specific per business :
- Depository function
- Fund administrator
- Fund Management
- Risk management
- Other …
- Statistical characteristics
- Links and dependencies
- Performance monitoring
- Dropping of related funds
- Lower signals
- Short- and medium-term forecasts
- Alert on tolerance exceedances
- Changes in liabilities
- Impact of the next significant redemption / subscription on the liquidity of the assets
- Characterization of the fund's liquidity
In addition to services related to the Big Data Platform, we support our clients in the areas of risk management, management support, fund administration, risk control and more ...
Our know-how resides in the resolution of financial problems by adapted quantitative models. We intervene on different subjects:
- Analysis of PnL
- Control and qualification of anonymous market data
- Problems of structure by term of rates, of CVA - DVA
- Model analysis: implied volatilities, pricing, risk, VaR, exchange rate
- Monitoring of prudential standards: Bâle 2, Bâle 3
- Allocation of assets
- Benchmarked management
- Development of proximity tools
For example for the control of the NAV, we use the date and the Mark To Market to get the low signals of the market from which, we calculate fractal indicators to obtain a fractal Benchmarck of the Fund, the economic indicators of the Fund, the fractal predictive value of the following NAV and the tolerance threshold. An alert is sent when the NAV estimated exceed the threshold.
The quality of indicators increases with the contribution of new information. Here are the back tests we did on some Funds.
The liabilities of a the Funds define its needs according to sinister past. Yet these disasters can evolve in an unexpected way. The low signals of the liabilities allow to ancitiper so the not contractual needs for the liabilities and to measure this impact for the management of the portfolio.
The exchanges are based on the request and the potential offer: the Security, the volume, the price, the matching (confrontation enters supply and demand). The trading has diverse origin: the effect at the end of accounting month, change of Security due to an under-performance, a result of a company, an announcement of number of unemployed, inflation, change of intervention rate of the central Bank and its consequences etc. … This change of behavior is appropriate to every agent and he does not respect a preestablished law.
The fractal methodology looks for the specific low signals (macroeconomic and other) in every Security and which can impact one the Mark To Market.
So the Bid/Ask spread can be impacted. This impact can also gaunt has character automobile-maintained by the system, the macroeconomic shocks, the rationality of the economic
golden agents another delay of passage of order. The number of record is very importing, over several hundred thousands.
In every case, the market by changing opinion on a price further to a new information, received a signal. It is exactly this not observable signal says low signal which the fractal methodology models. The objective being to anticipate the trend of the liquidity, its behavior with regard to the Security risky and its amplitude: overvalue, discount.
Fractal allows to study the region of stability or fork and the noise carried by the cost of the liquidity.
It is question of studying the functions of Bid and Ask in their own reference table about or the typology of the classes of Securities : sectorial, geographical or other. The analysis of the daily series shows the existence of a not stationarity, a random character, effects in the daytime (on Monday and Friday) bound to the closure of the quotation the weekend, effects months bound to the detachment of coupon of the securities and generally, impacts of business cycles. All these effects are not in the same cycle and are not either the same size.
The behavioral aspects (like American presidential cycle) and of the seasonality of financial markets constitute a not insignificant help to the decision-making regarding investment on Stock markets and especially in the weighting of the business sectors.
The identification of the low signals and the calculation of fractals indicators associated allows to improve the quality of the coefficients of weightings in particular the help to the construction of Asset allocation and follow-up of the risks of the portfolio.
Unlike Big Data, we use a limited number of data to obtain relevant information. However, these data will have to present big similarities with the gross series to be studied.
The Security given in guarantee can change quality following the low signals of the market. We look for fractals coefficients which qualify this liquidity of the Collateral. The exploration of fractale dimension is a runway among so many others.
One of the added values of the fractal model is the possibility of integrating into the methodology the know-how of the operational and the implementation of a threshold of overtaking according to the habits of this one. So, the decision-making tool alerts only when the tolerance thresholds are exceeded.