際際滷

際際滷Share a Scribd company logo

Presentation continuous series

Amit Sinha
Amit Sinha

This document discusses different methods for constructing continuous futures price series from discrete contracts for analysis and back-testing of trading strategies. It describes constant month series, constant distance/rolling nearby series, time to maturity weighted series, and volume/OI weighted series. The document provides an example of how a time to maturity weighted series is calculated and shows it produces a smoother price representation compared to a nearby series. It also compares the return distributions and strengths/weaknesses of nearby, weighted, and generic Bloomberg series.

1 of 11
Download to read offline
Continuous Future Series Construction Quant Team Ruchi Agri-Trading Singapore May 03, 2013
Why Continuous Future Series? We require a continuous historical price series for analysis or back-testing of trading strategies. One difficulty with the futures historical data is associated with splicing contracts together at contract boundaries. The results of any analysis will depend on the method we choose to construct the data series.
Different types of Continuous Series Constant Month Series (S Nov1, etc.) Constant Distance / Rolling Nearby / N-th Nearest Contract Series (S 1, BO2, etc.) Time to maturity weighted Series (Risk Metrics linear interpolation approach) Volume/OI weighted Series
Next Nearby/Rolling Nearby 13.6 13.8 14 14.2 14.4 14.6 14.8 15 15.2 Next Nearby Series S.X2 S.F3 Next.Nearby
Time to Maturity Weighted Series (RiskMetrics Method)
An Example: Time to Maturity Weighted Series Dates Contract Expiration Number of Days to Expiration Weights (%) Actual Contract Prices CMF Price NB1 1m CMF NB2 NB1 1m CMF NB2 爍燹材｀爻 爍燹材｀爻 燹酌ゞ燿燿 燹酌´燿燿 燹酌／燿燿 燹酌燿燿 2011- 07-11 07-14 08-10 08-12 3 30 32 0.10 0.90 13.5250 13.4650 13.4450 13.4710 2011- 07-12 07-14 08-11 08-12 2 30 31 0.07 0.93 13.6650 13.5550 13.5500 13.5623 2011- 07-13 07-14 08-12 08-12 1 30 30 0.03 0.97 13.8700 13.7475 13.7475 13.7516 2011- 07-14 07-14 08-15 08-12 0 30 29 0.00 1.00 13.9050 13.8200 13.7975 13.8200 2011- 07-15 08-12 08-15 09-14 28 30 61 0.93 0.07 - 13.8575 13.8300 13.8557 2011- 07-18 08-12 08-17 09-14 25 30 58 0.83 0.17 - 13.8550 13.8225 13.8496
Time to Maturity Weighted Series 13.5 14 14.5 15 15.5 16 Constant Distance Weighted Series S.X2 S.F3 Weighted
Comparison - Different Methods 0 20000 40000 60000 80000 100000 120000 140000 7 8 9 10 11 12 13 14 15 16 1/8/2008 1/9/2008 1/10/2008 1/11/2008 1/12/2008 Comparion b/w Weighted/Unweighted/Generic Bloomberg Series Nearby(Actual) Weighted Generic S1 VOLUME
Comparison - Different Methods -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 2/1/2008 2/3/2008 2/5/2008 2/7/2008 2/9/2008 2/11/2008 Difference b/w Weighted Series & Generic BBG Series BBG Generic - Weighted
Comparison - Return Distributions Next Nearby Weighted Std Dev 0.018 0.016 Skewness -0.905 -0.331 Kurtosis 20.612 4.860
Comparison - Different Methods Method Strengths Weaknesses Nth Nearest contract No data adjustment. Represents actual values for todays market. Gaps in historical data. Weighted Series Smoothens in data series. Good for statistical research. Non-tradable price representations. Constant Month Series Offers a long period of time in historical data on an annualized basis without adjusting the data. Small Open Interest/volume periods

More Related Content

Presentation continuous series

  • 1. Continuous Future Series Construction Quant Team Ruchi Agri-Trading Singapore May 03, 2013
  • 2. Why Continuous Future Series? We require a continuous historical price series for analysis or back-testing of trading strategies. One difficulty with the futures historical data is associated with splicing contracts together at contract boundaries. The results of any analysis will depend on the method we choose to construct the data series.
  • 3. Different types of Continuous Series Constant Month Series (S Nov1, etc.) Constant Distance / Rolling Nearby / N-th Nearest Contract Series (S 1, BO2, etc.) Time to maturity weighted Series (Risk Metrics linear interpolation approach) Volume/OI weighted Series
  • 5. Time to Maturity Weighted Series (RiskMetrics Method)
  • 6. An Example: Time to Maturity Weighted Series Dates Contract Expiration Number of Days to Expiration Weights (%) Actual Contract Prices CMF Price NB1 1m CMF NB2 NB1 1m CMF NB2 爍燹材｀爻 爍燹材｀爻 燹酌ゞ燿燿 燹酌´燿燿 燹酌／燿燿 燹酌燿燿 2011- 07-11 07-14 08-10 08-12 3 30 32 0.10 0.90 13.5250 13.4650 13.4450 13.4710 2011- 07-12 07-14 08-11 08-12 2 30 31 0.07 0.93 13.6650 13.5550 13.5500 13.5623 2011- 07-13 07-14 08-12 08-12 1 30 30 0.03 0.97 13.8700 13.7475 13.7475 13.7516 2011- 07-14 07-14 08-15 08-12 0 30 29 0.00 1.00 13.9050 13.8200 13.7975 13.8200 2011- 07-15 08-12 08-15 09-14 28 30 61 0.93 0.07 - 13.8575 13.8300 13.8557 2011- 07-18 08-12 08-17 09-14 25 30 58 0.83 0.17 - 13.8550 13.8225 13.8496
  • 7. Time to Maturity Weighted Series 13.5 14 14.5 15 15.5 16 Constant Distance Weighted Series S.X2 S.F3 Weighted
  • 8. Comparison - Different Methods 0 20000 40000 60000 80000 100000 120000 140000 7 8 9 10 11 12 13 14 15 16 1/8/2008 1/9/2008 1/10/2008 1/11/2008 1/12/2008 Comparion b/w Weighted/Unweighted/Generic Bloomberg Series Nearby(Actual) Weighted Generic S1 VOLUME
  • 9. Comparison - Different Methods -3 -2.5 -2 -1.5 -1 -0.5 0 0.5 1 2/1/2008 2/3/2008 2/5/2008 2/7/2008 2/9/2008 2/11/2008 Difference b/w Weighted Series & Generic BBG Series BBG Generic - Weighted
  • 10. Comparison - Return Distributions Next Nearby Weighted Std Dev 0.018 0.016 Skewness -0.905 -0.331 Kurtosis 20.612 4.860
  • 11. Comparison - Different Methods Method Strengths Weaknesses Nth Nearest contract No data adjustment. Represents actual values for todays market. Gaps in historical data. Weighted Series Smoothens in data series. Good for statistical research. Non-tradable price representations. Constant Month Series Offers a long period of time in historical data on an annualized basis without adjusting the data. Small Open Interest/volume periods