Let’s now see the accuracy of QM Portfolio Builder using the CSI data.
QM Portfolio Builder uses three separate data streams, Unadjusted, Dividend Adjusted and Split Adjusted. QM Portfolio Builder is designed to properly operate stock portfolios. In this case we will use the TLT ETF (iShares 20+ Year Treasury Bond).
We developed a “buy and hold” trading plan in which we reinvested dividends as paid on the TLT ETF. We will compare the results published on the IShares TLT site with our results extracted from QM Portfolio Builder.
IShares considers the purchase of “portions” of shares to be correct in the reinvestment of dividends. For example, it would be correct according to IShares to buy 2.3 shares of the “X” share. Obviously this is incorrect, as well as impossible. How does QM Portfolio Builder do it? If we have enough dividends to buy 2.4 shares, we will buy 2. If we have enough to buy 2.6 shares, we will buy 3. That is, rounding up or down will apply as appropriate. This is a reasonable and realistic estimate of the management of partial purchases.
Here we see the graph of IShares. We see that since this ETF began trading, the profitability obtained is 274.02%, reinvesting the dividends.
We are going to create a trading plan on QM Portfolio Builder with an account of 1,000,000 to minimize the effect of rounding on the purchase of shares. Another issue that we must consider is that we did not start operating this trading plan until August 8, 2002. During this period, the TLT ETF rose 91 basis points, this difference will be made worst by the effect of compound dividends on the extra share which starting earlier would of allowed them to buy.
The results obtained in QM Portfolio Builder using both rounding to reinvest the dividends and taking into account the 91 basis points that we missed, are astonishingly close to 1.8% error in 18 years! If we correct the start date, we drop to 89 basis points in the 18 years, or less than 5 basis points of error per year. A number surprisingly close to that obtained in IShares only by estimating the effect of partial share purchases and the effects of rounding in QM Portfolio Builder.
This is an example of matching returns using QM Portfolio Builder of one ETF. We did it in one ETF to make it easier to understand this analysis. Actually, QM Portfolio Builder was designed to work not only on a single ETF but on portfolios of stocks or ETF.
Most people use split or dividend adjusted data, which is invalid for portfolio analysis. Using Split and/or dividend adjusted data is only valid on a portfolio if you have the same dollar value of each stock and exit all positions at the same time. However, this is not realistic for trading systems. QM Portfolio Builder handles these problems and give you accurate results.
If you want more information, such as the trading session or the trading plan, you can send us an email to [email protected]
Article written by Quantified Models
Quantified Models YouTube Channel
We have several videos available on our YouTube channel that you may find very useful in developing trading systems.
We hope this information has been useful to you.
