Centralised stochastic simulation process

Abstract

A method of simulating a local system which is in interaction with other local systems within a global environment. The method includes generating, at a central location, a collection of global sets of values, each representing a possible future global state of the global environment at a future time, and the collection representing a variety of possible future global states. The method also includes transmitting the collection of global sets of values to the local systems. The local systems receive some of the collection, select a local set of parameters, and perform calculations on that data. The collection of values is used as a predictive simulation of the local system.

1 . A method of simulating a local system which is in interaction with other local systems within a global environment, wherein:
 said global environment has a global state that is represented by a global set of quantifiable entities,   the interaction of a local system with said global environment is represented at a given time by a local set of parameters, and   each local system has a predetermined local function to calculate a local indicator of performance, said predetermined local function using values of said global set of quantifiable entities and said local set of parameters,   said method comprising the steps of:   a) generating, at a central location separate from the local systems, a collection of global sets of values, each of said global sets of values representing a possible future global state of said global environment at a future time, and said collection representing a variety of possible future global states,   b) transmitting the collection of global sets of values to the local systems,   c) at one or more of the local systems,   c1) receiving at least some of said collection of global sets of values,   c2) selecting a first local set of parameters,   c3) calculating a first value of the local indicator of performance, with said predetermined local function using a first global set of values in said collection thereof as received, and using said selected first local set of parameters,   c4) selectively repeating step c3) while changing the global set of values within said collection thereof as received, until covering at least a portion of said collection, thereby obtaining a collection of values of the local indicator of performance, and   d) at the local systems, using said collection of values of the local indicator of performance as a predictive simulation of the local system, for said selected first local set of parameters thereof.   
     
     
         2 . The method of  claim 1 , wherein step c) further comprises:
 c5) selectively repeating steps c2) through c4) for another local set of parameters.   
     
     
         3 . The method of  claim 1 , wherein:
 step a) comprises generating said collection of global set of values for a plurality of horizons of time,   step b) comprises transmitting said collection of global set of values for said plurality of horizons of time, and   step c) is performed for at least a subset of said plurality of horizons of time.   
     
     
         4 . The method of  claim 1 , wherein step a) comprises:
 a1) at an expert site, remote from at least some of said local systems, building a statistical joint behavior model of a subset of values in said global set thereof, said statistical joint behavior model being described by its joint distribution density, and   a2) using said statistical joint behavior model to generate, at the central location, said collection of global set of values.   
     
     
         5 . The method of  claim 1 , wherein step a) comprises generating said collection of global set of values to represent the possible future states in terms of magnitude of the quantifiable entities. 
     
     
         6 . The method of  claim 1 , wherein:
 step a) comprises generating said collection of global sets of values to represent the possible future states in terms of variations of the quantifiable entities from a given condition thereof, and   step c1) comprises receiving said collection of global sets of values, and recalculating the collection of global sets of the quantifiable entities in terms of magnitude.   
     
     
         7 . The method of  claim 1 , wherein:
 step a) comprises generating said collection of global sets of values in the form of multi-valued samples representing possible future states of said global environment, and   the multi-valued samples comprise s*m*p possible numerical values for representing the possible future states, where p is the number of quantifiable entities in a set, s is the number of sets in the collection, and m is the number of horizons of time.   
     
     
         8 . The method of  claim 1 , wherein:
 step a) comprises generating said collection of global sets of values in the form of multi-valued samples representing possible future states of said global environment, and   the multi-valued samples comprise s*m*p possible numerical values for representing the change between possible future states and a reference state, for instance an observed current state, where p is the number of quantifiable entities in a set, s is the number of sets in the collection, and m is the number of horizons of time.   
     
     
         9 . The method of  claim 1 , wherein step b) comprises transmitting current state values together with said collection of global sets of numerical values. 
     
     
         10 . The method of  claim 1 , wherein
 step c) further comprises determining a statistical indicator related to a probability distribution of a future value of said local indicator of performance, from said collection of values thereof.   
     
     
         11 . The method of  claim 1 , wherein step c) further comprises:
 selectively repeating steps c2) to c4) for several sets of local parameters, for each local set of parameters,   i) determining one or more statistical quantities related to the probability distribution of a future value of said local indicator of performance, from said collection of values thereof, as a function of said local set of parameters,   ii) calculating the value of an optimization criterion, defined as a function of said one or more statistical quantities, the function being associated with the local set of parameters, and   iii) selecting an optimal local set of parameters that optimizes the optimization criterion.   
     
     
         12 . The method of  claim 11 , wherein step c) further comprises:
 iv) setting the local set of parameters to match said optimal local set of parameters.   
     
     
         13 . The method of  claim 1 ,
 wherein step a) further comprises:   a3) maintaining, at the central location, a performance evaluation function for each of a plurality of possible local parameters of action,   wherein step b) further comprises:   b1) transmitting such performance evaluation functions, each in association with a respective parameter identifier, and   wherein step c3) comprises calculating said first value of the local indicator of performance by components, with a component of the local indicator of performance being calculated through applying a transmitted performance evaluation function to the corresponding local parameter, as defined by the parameter identifier.   
     
     
         14 . The method of  claim 11 , wherein, in step c3), a component of the local indicator of performance for a local parameter is proportional to that local parameter. 
     
     
         15 . The method of  claim 1 , wherein step a) comprises deriving observed values from sensors of the environment, and deriving a current set of values from said observed values through a selected model, and
 step b) comprises transmitting the current set of values as part of said collection of global sets of values.   
     
     
         16 . The method of  claim 15 , wherein said selected model uses an historic storage of said observed values. 
     
     
         17 . The method of  claim 16 , wherein:
 step a) comprises generating in a centralized way said collection of global sets of values, using a Monte-Carlo method and a generator of pseudo random numbers.   
     
     
         18 . The method of  claim 16 , wherein:
 step a) comprises generating, at the central location, said collection of global sets of values, using an historical method based on a plurality of past sets of values, and a probability indicator for each such past sets of values.   
     
     
         19 . The method of  claim 16 , wherein step a) comprises:
 introducing statistical joint behavior biases in said collection of global sets of values.   
     
     
         20 . The method of  claim 19 , wherein step a) comprises:
 introducing said statistical joint behavior biases in said collection of global sets of values upon request of a local system.   
     
     
         21 . The method of  claim 19 , wherein step a) comprises:
 locally generating said statistic joint behavior biases in a local system.   
     
     
         22 . The method of  claim 1 , wherein steps a) and b) are performed substantially periodically. 
     
     
         23 . The method of  claim 1 , wherein said transmission at step b) is a broadcast transmission. 
     
     
         24 . The method of  claim 1 , wherein said transmission at step b) comprises peer-to-peer downstream transmissions that are sent to a plurality of local systems. 
     
     
         25 . The method of  claim 1 , wherein said transmission at step b) comprises peer-to-peer upstream transmissions that are obtained from a plurality of local systems. 
     
     
         26 . The method of  claim 3 , wherein said collection of global sets of values comprises a first level sub-file for each horizon of time, each first level sub-file in turn including a plurality of second level sub-files. 
     
     
         27 . The method of  claim 3 , wherein each horizon of time is associated with a time horizon indicator, the method further comprising creating a separate file or folder of files containing the corresponding subset of said global set of values for each horizon of time. 
     
     
         28 . The method of  claim 1 , further comprising creating a separate file or folder of files containing the corresponding subset of said global set of values for each a plurality of drawing indicators respectively associated with a plurality of drawing scenarios. 
     
     
         29 . The method of  claim 1 , wherein said transmission at step b) is a broadcast transmission, and step c) comprises using a locally pre-selected portion of the collection of global sets of values, as received. 
     
     
         30 . The method of  claim 1 , wherein said transmission at step b) comprises a plurality of point-to-point transmissions, each comprising a respective pre-selected portion of said collection of global sets of values. 
     
     
         31 . The method of  claim 1 , wherein said transmission at step b) comprises offering a peer to peer access to a local system, each peer to peer access comprising a respective pre-selected portion of said collection of global sets of values. 
     
     
         32 . The method of  claim 1 , wherein steps c2) to c4) are selectively repeated for different sets of local parameters. 
     
     
         33 . The method of  claim 32 , further comprising applying, by one or more local systems, an algorithm to find an optimal set of local parameters with respect to a criterion that is either common or specific to each local system and which can be computed from the said collection of values of the local indicator of performance. 
     
     
         34 . The method of  claim 32 , wherein step d) is followed by the following steps:
 at one or more of the local systems, retransmitting the local indicator of performance back to a remote location that is common to the subset of local systems,   at said remote location,
 i) receiving a plurality of local indicators of performance from said subset of the local systems, 
 ii) calculating a global indicator of performance, using a global evaluation model, 
 iii) comparing the global indicator of performance with the plurality of local indicators of performance, and determining a global setting parameter from said comparison, and 
   sending to each local system information derived from said global setting parameter, to utilize in calculating a value of the local indicator of performance at step c3).   
     
     
         35 . The method of  claim 32 , wherein:
 step iii) comprises determining a plurality of global setting parameter values, as a function of a plurality of local set of parameters, and   the sending step comprises sending to each local system reference data for determining said local set of parameters, said reference data being derived from optimizing the function of step iii).   
     
     
         36 . The method of  claim 1 , wherein the local systems include financial operators. 
     
     
         37 . The method of  claim 1 , wherein the global environment includes global financial markets. 
     
     
         38 . The method of  claim 1 , wherein the quantifiable entities include prices at which financial securities are traded, and wherein at least one of the local systems to be monitored and controlled is a financial market. 
     
     
         39 . The method of  claim 1 , wherein the local parameters include quantities of traded securities in a portfolio. 
     
     
         40 . The method of  claim 1 , wherein the local indicator of performance includes a valuation of a portfolio of at least one of the financial operators, and wherein at least one of the local systems to be monitored and controlled is a financial market. 
     
     
         41 . The method of  claim 1 , wherein the global set of values includes a set of market variables from which the price of all traded securities can be recreated with a predetermined level of precision, using a local assessment model. 
     
     
         42 . The method of  claim 10 , wherein the statistical indicator is at least one of the following:
 Value-at-Risk; and   Standard deviation.   
     
     
         43 . The method of  claim 1 , wherein step a) comprises introducing biased data within the collection of global sets of values, said biased data corresponding to a catastrophic scenario. 
     
     
         44 . The method of  claim 1 , wherein step a) comprises introducing biased data within the collection of global sets of values as received, said biased data corresponding to a catastrophic scenario specified by a request from a local system. 
     
     
         45 . The method of  claim 1 , wherein step c1) comprises introducing biased data within the collection of global sets of values as received, said biased data corresponding to a catastrophic scenario. 
     
     
         46 . A system for simulating a local system which is in interaction with other local systems within a global environment, wherein said global environment has a global state that is represented by a global set of quantifiable entities, the interaction of a local system with said global environment is represented at a given time by a local set of parameters, and each local system has a predetermined local function to calculate a local indicator of performance, said predetermined local function using values of said global set of quantifiable entities and said local set of parameters,
 said system comprising:   generating means, provided at a central location separate from the local systems, for generating a collection of global sets of values, each of said global sets of values representing a possible future global state of said global environment at a future time, and said collection representing a variety of possible future global states,   transmitting means for transmitting the collection of global sets of values to the local systems,   at one or more of the local systems:
 first receiving means for receiving at least some of said collection of global sets of values, 
 selecting means for selecting a first local set of parameters, 
 first calculating means for calculating a first value of the local indicator of performance, with said predetermined local function using a first global set of values in said collection thereof as received, and using said selected first local set of parameters, and 
 repeating means for selectively causing the first calculating means to be repeated while changing the global set of values within said collection thereof as received, until covering at least a portion of said collection, thereby obtaining a collection of values of the local indicator of performance, 
 using means for using said collection of values of the local indicator of performance as a predictive simulation of the local system, for said selected first local set of parameters thereof.