Need Help with coding in Python
timtimnbebe
csc559_-__price_generator_-_2014___3.pptby
Mark Goetsch
*
The Pieces
Arbitrage
Quotes
Order Matching
STP
*
*
*
*
“I enter orders[1] in an order book[2] for a specific market[3]. The orders are matched at discrete time time intervals that are controlled by a heartbeat[4]. The orders can be either buy[5] or sell[6] orders[1]. The order matcher will will check whenever a new order[1] is entered into the market. The time will be marked by the heartbeat[4]. If there is a match[7] then the order matcher will either fully match the order[8] or partially match the order[9].”
Patterns Used
Orders
Orderbook
Market
Heartbeat
Buy Order
Sell Order
Match
Full Match
Partial Match
*
*
*
Time Beat
Invalid Order
Match
Match Rules
Sell Orders
Buy Orders
Full Match
Partial Match
Orders
*
*
*
A customer authenticates[1] to a particular contract[2] which belongs to a market[3]. An order[5] is then presented to the orderbook[4] but not before checking the order[5] against the customers margin[6] which is different for every contract[2].
Gateway Analysis Pattern
Patterns Used
Authenticate
Contract
Market
Orderbook
Order
Margin
*
*
*
Authentication
Margin
Contract
Order
Market
Order Book
*
*
*
Rules that are used for every matching possibility.
*
*
*
*
*
Opening Rules =>
How are orders entered before the opening
Closing Rules =>
How are orders handled at the end
Trading Rules =>
How are orders matched
*
*
*
Limit Order
Against
Limit Order
Buy Limit in book >= incoming sell Limit
Sell Limit in book <= incoming buy Limit
Match buy quantity
Best Buy Limit
Match sell quantity
Best Sell Limit
“The incoming order is a limit order. It matches against the best limit in the book (bid price >= ask price) for the incoming limit order. The best limit in the book determines the price”
Remaining buy quantity
Remaining sell quantity
*
*
*
Three Stages to Opening the Market
Staging Period – Orders are entered and a countdown begins which calculates the Indicative Opening Price (IOP) is possible (see table below).
Non-Cancel Period – Can enter new orders but not cancel orders. The final IOP is calculated.
Opening – Orders that can be matched are matched.
| Pre-Opening State of Book | Conditions | Settle Exists | No Settle Exists |
| No Entries | --nothing-- | Settle | --nothing-- |
| Bids Only | Bid > Settle Bid = Settle Bid < Settle | Bid Bid Settle | Bid Bid Bid |
| Offers Only | Offer > Settle Offer = Settle Offer < Settle | Settle Offer Offer | Offer Offer Offer |
| Bids and Offers No Trades Possible | Settle = Offer Settle = Offer Bid < Settle < Offer Bid = Settle Bid > Settle | Offer Offer Settle Bid Bid | Bid Bid Bid Bid Bid |
| Bids and Offers Trades Possible | Bid > Offer | IOP (Indicative Opening Price) | IOP (Indicative Opening Price) |
*
*
*
*
Order Matcher
Front End Connectivity
CME GLOBEX API
CBOEdirect API
CBOE Trade Match
CME GLOBEX
Trade Match
Trade Processing and Clearing
*
*
CMI
FIX 4.2
Ilink FIX 4.2
FIX 4.2
FIX 2.3 Express
Confirm Record
CMTA/
Allocation
Trade Correction
Confirm Record
GUS/
Allocation
Trade Correction
Firm Back Office Systems
FCM Back Office Systems
OCC
CME Clearing
*
*
*
*
95.00
95.05
95.10
95.15
95.20
95.25
Orders
Price Generator
Timer++
Limit Orders
Add Order
[Array List]
Match
[FIFO Doubly Linked List]
Matched Orders
[Queue]
[Queue]
Knowing financial theory
How to Test the Engine
Testing Engine
Trading Engine
Sample Buy and Sell Transactions
Various prices above and below the book are generated according to a random distribution.
https://www.khanacademy.org/economics-finance-domain/core-finance/derivative-securities/Black-Scholes/v/introduction-to-the-black-scholes-formula
https://www.khanacademy.org/economics-finance-domain/core-finance/derivative-securities/forward-futures-contracts/v/motivation-for-the-futures-exchange
Random Walk
public static double SimulateAsset(double s0, double mu, double sigma,double tau, double delta_t,MCG g)
{
//Purpose: Simulates an Asset Price run using a random walk and returns a final asset price.
// so = Price of the asset at time 0 (current time)
// mu = Historical Mean
// sigma = Historical Volatility (variance)
// delta_t = period of time (% of a year or a day)
// g = Random variable
double s = s0;
// Made the steps = to the number of days which is the same as daily changes.
double nSteps =tau;
for (int i=0; i < (int)nSteps; i++)
{
// s = s0 * (1 + mean + standard deviation * gaussian random number * squareRoot of the time period.
s= s * (1 + mu * delta_t + sigma * g.gaussian() * Math.sqrt(delta_t));
}
//Returns the final Price
return s;
}
Simulating an Asset as a Random Walk (or drunkards walk)
*
public static double MeasureVolatilityFromHistoric(double[] historic, double delta_t, int length)
{
// Purpose: Measures the Volatility for scaled prices.
double sum = 0;
double variance = 0;
double volatility = 0;
// length - 1 instead of length since n prices generates n-1 returns
for (int i=0; i< length -1; i++)
{
//Random variable X^2
sum = sum + Math.pow((historic[i+1]-historic[i])/historic[i],2);
}
// E[X^2] - E[X]^2
variance =
sum / (length -1) - Math.pow(MeasureMeanFromHistoric( historic, delta_t,length) * delta_t,2);
// Volatility = SquareRoot(variance/ dt) which is the standard deviation scaled for a time increment
volatility = Math.sqrt(variance/delta_t);
return volatility;
}
Measuring the Volatility Associated with the Trade
*
public static double MeasureMeanFromHistoric(double[] historic, double delta_t, int length)
{
//Purpose: Measures the mean of the scaled prices. (Scaled indicates that the level of the
// Prices is not important.
double sum = 0;
double average = 0;
double waverage = 0;
double returns = 0;
//length-1 because the scaling requires n prices to generate a sequence of n-1 scaled returns.
for (int i=0;i < (length-1); i++)
{
// Scales the returns and sums them
returns = (historic[i+1]-historic[i])/historic[i];
sum = sum + returns;
}
//computes the average of the returns
average = sum/(length-1);
// divides the average by dt so that the average applies to each time increment
waverage = average/delta_t;
return waverage;
}
Measuring the averages
*
Defining your engine
Patterns Used
Quote
Instrument, bid, offer, number, spread (bid-offer), mid (bid+offer / 2), one-way quote, two-way quote.
Board Exercise
“To understand the value of a contract, we need to understand the price of the goods being traded. Goods are often priced differently depending on whether they are bought or sold. This two-way pricing behavior can be captured by a quote[1]”
Martin Fowler, Analysis Patterns 1997
*
Patterns Used
Scenario
Instrument, Quote, Timepoint, Price, Quote, Party, Information Source, Market Indicator
Board Exercise
“In volatile markets, prices can change rapidly. Traders need to value goods against a range of possible changes. The scenario[1] puts together a combination of conditions that can act as a single state for the market for valuation. Scenarios can be complex, and we need a way to define their construction so we can use the same scenario construction at different times in a consistent manner. Scenarios are useful for any domain with complex price changes.”
Martin Fowler, Analysis Patterns 1997
*
ud Use Case Model
OrderMatcher
Order
Enter an
Order
Delete Order
Modify Order
Check Order
Validate
Order
Order
Matched
Clearing System
Account System
Margin System
«include»
«include»
«include»
«include»
«include»
95.0095.0595.1095.1595.2095.25
Orders
Price
Generator
Timer++
Limit Orders
Add OrderMatch
Matched
Orders
[Array List]
[FIFO Doubly Linked List]
[Queue]
[Queue]
