Financial Data

An example that demonstrates some of ojAlgo's features related to working with financial data. The example shows how to:

1. Download historical data for Apple and Google stock from Yahoo Finance.
2. Synchronize/coordinate the two series so that they have the same length, beginning and end, as well as resample them to be monthly data.
3. Statistics are compared for Apple and Google stock, and the correlation between them is calculated.
4. Use that data to estimate (the parameters of) stochastic processes. For each time series two processes are created - one with a monthly scale and one annualized.
5. For Apple stock the two generated processes are verified to produce the same probability distributions, repectively, at one month and one year in to the future.
6. The probability distributions, one year into the future, of Apple and Google are compared.
7. The stochastic process for Google is used to simulate a 1000 future scenarios, and the scenmarios' value one year into the future is compared to theoretical distribution.

Example code

import java.util.Map.Entry;

import org.ojalgo.OjAlgoUtils;
import org.ojalgo.finance.data.YahooSymbol; //need ojalgo-finance for this
import org.ojalgo.netio.BasicLogger;
import org.ojalgo.random.SampleSet;
import org.ojalgo.random.process.GeometricBrownianMotion;
import org.ojalgo.random.process.RandomProcess.SimulationResults;
import org.ojalgo.series.CalendarDateSeries;
import org.ojalgo.series.CoordinationSet;
import org.ojalgo.series.primitive.DataSeries;
import org.ojalgo.type.CalendarDateUnit;

public abstract class FinancialData {

    public static void main(final String[] args) {

        BasicLogger.debug();
        BasicLogger.debug(FinancialData.class.getSimpleName());
        BasicLogger.debug(OjAlgoUtils.getTitle());
        BasicLogger.debug(OjAlgoUtils.getDate());
        BasicLogger.debug();

        final YahooSymbol tmpAAPL = new YahooSymbol("AAPL", CalendarDateUnit.DAY);
        final YahooSymbol tmpGOOG = new YahooSymbol("GOOG", CalendarDateUnit.DAY);

        final CalendarDateSeries<Double> tmpSeriesAAPL = tmpAAPL.getPriceSeries();
        final CalendarDateSeries<Double> tmpSeriesGOOG = tmpGOOG.getPriceSeries();

        final CoordinationSet<Double> tmpUncoordinated = new CoordinationSet<>();
        tmpUncoordinated.put("Apple", tmpSeriesAAPL);
        tmpUncoordinated.put("Google", tmpSeriesGOOG);

        BasicLogger.debug("Downloaded daily data");
        for (final Entry<String, CalendarDateSeries<Double>> tmpEntry : tmpUncoordinated.entrySet()) {
            BasicLogger.debug("\tSeries: {} = {}", tmpEntry.getKey(), tmpEntry.getValue());
        }

        final CoordinationSet<Double> tmpCoordinated = tmpUncoordinated.prune(CalendarDateUnit.MONTH);

        BasicLogger.debug();
        BasicLogger.debug("Coordinated monthly data");
        for (final Entry<String, CalendarDateSeries<Double>> tmpEntry : tmpCoordinated.entrySet()) {
            BasicLogger.debug("\tSeries: {} = {}", tmpEntry.getKey(), tmpEntry.getValue());
        }

        final DataSeries tmpMonthlyDataAAPL = tmpCoordinated.get("Apple").getDataSeries();
        final DataSeries tmpMonthlyDataGoogle = tmpCoordinated.get("Google").getDataSeries();

        final SampleSet tmpYahooSet = SampleSet.wrap(tmpMonthlyDataAAPL.log().differences());
        final SampleSet tmpGoolgleSet = SampleSet.wrap(tmpMonthlyDataGoogle.log().differences());

        BasicLogger.debug();
        BasicLogger.debug("Sample statistics (logarithmic differences on monthly data)");
        BasicLogger.debug("\tApple:  {}", tmpYahooSet);
        BasicLogger.debug("\tGoogle: {}", tmpGoolgleSet);
        BasicLogger.debug("\tCorrelation: {}", tmpGoolgleSet.getCorrelation(tmpYahooSet));

        final GeometricBrownianMotion tmpAppleProcMonth = GeometricBrownianMotion.estimate(tmpMonthlyDataAAPL, 1.0);
        final GeometricBrownianMotion tmpGoogleProcMonth = GeometricBrownianMotion.estimate(tmpMonthlyDataGoogle, 1.0);
        tmpAppleProcMonth.setValue(1.0); // To normalize the current value
        tmpGoogleProcMonth.setValue(1.0);

        final double tmpYearsPerMonth = CalendarDateUnit.YEAR.convert(tmpCoordinated.getResolution());

        final GeometricBrownianMotion tmpAppleProcYear = GeometricBrownianMotion.estimate(tmpMonthlyDataAAPL, tmpYearsPerMonth);
        final GeometricBrownianMotion tmpGoogleProcYear = GeometricBrownianMotion.estimate(tmpMonthlyDataGoogle, tmpYearsPerMonth);
        tmpAppleProcYear.setValue(1.0); // To normalize the current value
        tmpGoogleProcYear.setValue(1.0);

        // Comparing the month and year based stochastic processes for AAPL

        BasicLogger.debug();
        BasicLogger.debug("    Apple \t Monthly proc \t\t\t Annual proc \t\t (6 months from now)");
        BasicLogger.debug("Expected: {} \t {}", tmpAppleProcMonth.getDistribution(6.0).getExpected(), tmpAppleProcYear.getDistribution(0.5).getExpected());
        BasicLogger.debug("StdDev:   {} \t {}", tmpAppleProcMonth.getDistribution(6.0).getStandardDeviation(),
                tmpAppleProcYear.getDistribution(0.5).getStandardDeviation());
        BasicLogger.debug("Var:      {} \t {}", tmpAppleProcMonth.getDistribution(6.0).getVariance(), tmpAppleProcYear.getDistribution(0.5).getVariance());

        // Comparing the annualized stochastic processes for AAPL and GOOG

        BasicLogger.debug();
        BasicLogger.debug(" \t\t\t Apple \t\t\t\t\t Google");
        BasicLogger.debug("Current:  {} \t\t\t\t\t {}", tmpAppleProcYear.getValue(), tmpGoogleProcYear.getValue());
        // getExpected() is a shortcut to getDistribution(1.0).getExpected()
        BasicLogger.debug("Expected: {} \t {}", tmpAppleProcYear.getExpected(), tmpGoogleProcYear.getExpected());
        // getStandardDeviation() is a shortcut to getDistribution(1.0).getStandardDeviation()
        BasicLogger.debug("StdDev:   {} \t {}", tmpAppleProcYear.getStandardDeviation(), tmpGoogleProcYear.getStandardDeviation());
        // getVariance() is a shortcut to getDistribution(1.0).getVariance()
        BasicLogger.debug("Var:      {} \t {}", tmpAppleProcYear.getVariance(), tmpGoogleProcYear.getVariance());

        final SimulationResults tmpSimResults = tmpGoogleProcYear.simulate(1000, 12, tmpYearsPerMonth);
        BasicLogger.debug();
        BasicLogger.debug("Simulate future Google");
        BasicLogger.debug("Simulated sample set:  {}", tmpSimResults.getSampleSet(11));

    }

}

Console output

FinancialData
ojAlgo
2016-05-14

Downloaded daily data
    Series: Google = GOOG DAY First:2004-08-19 12:00:00=50.119968 Last:2016-05-13 12:00:00=710.830017 Size:2955
    Series: Apple = AAPL DAY First:1980-12-12 12:00:00=0.431358 Last:2016-05-13 12:00:00=90.519997 Size:8932

Coordinated monthly data
    Series: Google = GOOG MONTH First:2004-08-31 12:00:00=51.133953 Last:2016-05-31 12:00:00=710.830017 Size:142
    Series: Apple = AAPL MONTH First:2004-08-31 12:00:00=2.26768 Last:2016-05-31 12:00:00=90.519997 Size:142

Sample statistics (logarithmic differences on monthly data)
    Apple:  Sample set Size=141, Mean=0.02614761351196476, Median=0.033231314721613, Var=0.01049149860307574, StdDev=0.10242801669014069, Min=-0.39981831885234254, Max=0.30177586998374273
    Google: Sample set Size=141, Mean=0.01866655749244666, Median=0.01752020055681669, Var=0.008705486904812838, StdDev=0.09330319879196446, Min=-0.20428855960758607, Max=0.3859340135368754
    Correlation: 0.492084812100738

    Apple    Monthly proc            Annual proc         (6 months from now)
Expected: 1.2072682671701467     1.2072682671701467
StdDev:   0.307729141701079      0.30772914170107896
Var:      0.09469722465208277    0.09469722465208275

             Apple                   Google
Current:  1.0                    1.0
Expected: 1.4574966689160085     1.3181531278846041
StdDev:   0.5338626633590423     0.43741531662489624
Var:      0.28500934332881017    0.1913321592180582

Simulate future Google
Simulated sample set:  Sample set Size=1000, Mean=1.325627179042084, Median=1.2687505025091892, Var=0.1856145903552972, StdDev=0.43083011774398644, Min=0.4581958805895357, Max=3.6433282195001504