Performance
Pair trading is one of the most reliable strategies of professional traders.
Trading stock pairs help to hedge sector- and market-risk. For example, if the whole market crashes and the two stocks plummet along with it, the trade should result in a gain on the short position and a negating loss on the long position. Despite the significant market movement, our pair profit remains neutral and we can further benefit from the expected return of the price ratio to the normal.
Analysis of 324,000 pairs of 132 sectors of the US economy showed that 65% of pairs are profitable - without any optimization of the model parameters!
Trading many (tens or hundreds) stock pairs brings a high degree of probability that the portfolio will be profitable every year - see below.
Due to limited and well-controlled risk, diversified portfolio and statistical advantage the strategy generate stable profits for several decades.
Overall statistics
The following graphs summarizes an extensive study of the profitability of stock pairs. 324,125 stock pairs of 132 sectors of the US economy was tested using
Stock Pair Builder . Pairs were tested on the model Ratio with default settings (15 to 15 - 2.0 to 0.0) for a period of 1, 2, 3 and 4 years.
Portfolio profitability
The following table shows the percentage of profitable portfolios depending on the number of pairs and the test period. Note that the more pairs is in the portfolio, the greater the likelihood that it will be profitable. It is a consequence of the high degree of profitability of pairs (65%). We see that trading 50 randomly selected pairs gives us 95% confidence that the portfolio will be profitable in first year.
Testing Methodology:
Of all 324,125 of potential stock pairs available in defined sectors (132) we have chosen 10x 10x 100 = 10,000 random portfolios. Each portfolio contained N (5-50) pairs.
Portfolios have been tested over the last 1, 2, 3 ... 10 years. It means that we have tested 100 random portfolios for each combination NumberOfPairs x Period.
The number in the table shows number of profitable portfolios with given number of pairs (table row) on given period (table column).
Performance of portfolio "Building materials"
Fig. 3: Real portfolio performance in 2006 - 2016 (commissions 4$/RT)
Portfolio built from stocks in sector "Building Materials": CSTE, CPAC, CX, CCCL, CBPX, CUO, CPSH, CRH, EXP, FRTA, HW, JHX, USCR, USG, YECO
Performance of portfolio "Electronic Components"
Fig. 4: Real portfolio performance in 2006 - 2016 (commissions 4$/RT)
Portfolio built from stocks in sector "Electronic Components": GME, DRAM, IMN, MSDI, MSDIW, NTAP, NMBL, PSTG, QBAK, QTM, STX, WDC, ARW, AVT, CUI, IGC, IFON, ITRN, RADA, RELL, TAIT, TEL, TESS, UUU, VOXX, TGLS, CAMT, CYBE, IIVI, KLAC, NVMI, PRCP
Performance of portfolio "Computer manufacturing"
Fig. 5: Real portfolio performance in 2006 - 2016 (commissions 4$/RT)
Portfolio built from stocks in sector "Computer manufacturing": AAPL, CCUR, CRAY, DVMT, FIT, HPE, HPQ, IBM, LXK, NICE, OMCL, SCKT, SMCI, TDC, XPLR
Performance of portfolio "Packed food"
Fig. 6: Real portfolio performance in 2006 - 2016 (commissions 4$/RT)
Portfolio built from stocks in sector "Packed food": APFH, ALN, ANFI, BETR, ADM, BGS, BUFF, BG, BUR, CPB, CAG, DAR, DF, FARM, FLO, FRPT, GIS, TWNK, TWNKW, INGR, SNAK, SJM, K, LANC, LWAY, MKC, MKC.V, MJN, MED, MDLZ, NOMD, OME, PF, POST, RIBT, RIBTW, SENEA, SENEB, SPU, SYUT, HAIN, KHC, THS, WWAV
Incredible? Download Stock Pair Builder and test it yourself.
All above mentioned sectors are predefined in the Builder. You can see results in 2 minutes.