I was curious how lizgdal's timing measure [SMA(1,42) / SMA(211,42) ratio being less than 0.9] worked when we had an ETF to invest in the Nasdaq 100. QQQ started on March 10, 1999 (at least that's what Yahoo's historical data shows), so I looked at a few things starting then. (I didn't realize this before, but Robbie's backtester has the QQQ history available, unlike many other stock or ETF symbols.)

First, it appears that Robbie's N1T models QQQ pretty well (as I'm sure Robbie showed long ago).

QQQ: gtr1.net: GTR1 Backtester

N1T: gtr1.net: GTR1 Backtester

3/10/99 - 11/28/25

                   QQQ    N1T
                  ----   ----
CAGR              10.5   10.3
GSD(20)           27.3   26.6
Sharpe(20)        0.46   0.46
SAWR(20; 0.95)     4.0    3.9
LDDD3             19.6   18.8
MDD              -83.0  -81.1

Next, I looked to see if QQQ results improved if this timing method was used for the period starting 3/10/1999. The link below changes the first step to "dspo(1) >= 252" to "... >= 0". Without this, the backtest has you in cash for the first year. If you backtest N1T during this period, there is prior history so it has you invested in that first year.

QQQ with SMA timing: gtr1.net: GTR1 Backtester

3/10/99 - 11/28/25
                            QQQ_SMA
                   QQQ   (42,211,0.9)
                  ----   ------------
CAGR              10.5       15.0
GSD(20)           27.3       19.9
Sharpe(20)        0.46       0.76
SAWR(20; 0.95)     4.0       10.3
LDDD3             19.6        9.6
MDD              -83.0      -46.6

So, improvement in both return and risk metrics.

I was curious about when this helped and didn't help, so I simply divided the 27-plus year history into halves, with the first half ending on 7/17/2012.

3/10/99 - 7/17/12
                            QQQ_SMA
                   QQQ   (42,211,0.9)
                  ----   ------------
CAGR               2.0       10.4
GSD(20)           33.4       21.3
Sharpe(20)        0.13       0.49
SAWR(20; 0.95)     0.9        6.4
LDDD3             26.6       11.8
MDD              -83.0      -46.6

7/18/12 - 11/28/25
                            QQQ_SMA
                   QQQ   (42,211,0.9)
                  ----   ------------
CAGR              19.5       19.6
GSD(20)           19.7       18.3
Sharpe(20)        1.00       1.07
SAWR(20; 0.95)    14.4       14.7
LDDD3              7.9        6.5
MDD              -35.1      -28.7

Two very different investing "eras". All the benefit of the timing occurs in the first half, with the second half showing no or very little improvement in return/risk metrics. No real surprise, given the nature of the markets during these two periods (March 1999 was just a year away from the March 2000 peak; there were longer bear markets in the first half; and the second half had much higher untimed returns, so it was harder for timing to provide any help).

I also tried to find more optimal parameters in each half. Although I could find some different parameters that yielded better results, the improvements were minimal. The SMA (42, 211, 0.9) parameters seem pretty robust for these two periods.

Will this timing method be useful in the near future? Depends on what type of investing era we think we're in.