Author: rgearyiii
Subject: The GTR1 Hacktester
Date: 10/30/2019
"... Nas100Momentum, as defined in the sample data, refers to a simple screen that buys the top 5 NASDAQ 100 stocks by one-year total return and holds for 252 market days. ..."
gtr1.net: GTR1 Backtester - Nas100MomentumAnnual_20191030_rgearyiii
reposted from https://yorickm.com/Message.php?pid=34330952

ProShares Nasdaq-100 Dorsey Wright Momentum ETF QQQA inception date May 18, 2021.

Author: borisnand
Subject: Re: The GTR1 Hacktester
Date: 11/7/2021
... This screen wasn`t originally posted as anything special, but the screen`s extremely high returns in 2020 might make it useful as a benchmark for other Large Growth Momentum screens.

Screen: Buy 5 Nasdaq-100 stocks with the highest 1-year total return. Monthly trading improves results significantly in a backtest from 1985 to 2021.

37-year backtest from 19850201 to 20211105

      Screen       CAGR  GSD  MDD  Sharpe  Beta  AT
  Nas100Momentum    17   45   -91   0.54   1.5   0.9  annual
  Nas100Momentum    29   42   -74   0.81   1.4   4.2  monthly
SP1500EqualWeight   13   22   -60   0.57   1.1   0.4

15-year backtest from 20061106 to 20211105

      Screen       CAGR  GSD  MDD  Sharpe  Beta  AT
  Nas100Momentum    19   36   -68   0.69   1.3   0.9  annual
  Nas100Momentum    24   34   -54   0.87   1.1    4   monthly
SP1500EqualWeight   11   26   -60   0.56   1.2   0.3

{Nas100MomentumMonthly} had a blowout year in 2020, outperforming the benchmark by over 100%, but is trailing in 2021.

year  Nas100Mo  SP1500ew  excessR
1985     40        31        9
1986     28        18       10
1987     49         5       45
1988     42        20       22
1989     72        26       46
1990      9       -12       21
1991    157        35       121
1992      8        15       -8
1993     20        15        5
1994     16         1       15
1995     31        28        3
1996     16        21       -4
1997     32        27        4
1998    104         6       99
1999    186        11       175
2000    -17        14       -31
2001    -19        10       -29
2002    -25       -15       -10
2003     66        44       21
2004     25        19        6
2005     36         8       28
2006      5        16       -11
2007     44        -1       45
2008    -42       -36       -6
2009     10        46       -36
2010     32        26        6
2011     -2        -1        0
2012     23        18        5
2013     55        39       16
2014     38         9       29
2015     22        -5       27
2016     25        24        2
2017     39        14       25
2018      1       -10       11
2019     33        25        9
2020    119        14       105
2021     23        33       -10

data from gtr1.net: GTR1 Backtester - Nas100MomentumMonthly_20191030_rgearyiii

The worst 5-year excess return was February 2000 to 2005, with Nas100Mo -6% CAGR and SP1500ew 14% CAGR. This underperformance was rare though. In the backtest, 89% of 5-year excess returns for {Nas100MomentumMonthly} were positive. The extremely high returns in 1999 make the use of rolling returns useless for timing when to stop using the screen. BCC timing (to get out) combined with rolling returns (to get in) might work.

There`s a new ETF using a similar idea. ProShares Nasdaq-100 Dorsey Wright Momentum ETF ($QQQA) picks 21 Nasdaq-100 stocks based on a proprietary momentum signal (based on historical point and figure technical analysis), rebalances to equal weight every quarter, and uses htd to reduce turnover.

I don`t think I would use this screen, but it`s a good benchmark for other Large Growth Momentum screens. I don`t want to compete directly against an ETF, and the idea is simple enough that there are probably other competitors out there tweaking their momentum measures. I`ll fish where I see fewer fishermen.
reposted from https://yorickm.com/Message.php?pid=34973261