Summary: No you will not.

Background

What is the NYSI signal?

Popularized on Twitter by popular fintwit users such as OptionsHawk, ukarlewitz, and others, an upcross of the NYSI (Mcclellan Summation Index) over the 8EMA indicates an uptrend in the broader market. And vice versa – a downcross of the NYSI under the 8EMA indicates a downtrend in the broader market.

However, it seems to just serve as a general purpose guide, and cannot be acted upon in a way that produces alpha.

Results

Since inception of the NYSI signal on Stockcharts.com in June 1998:

Following the Up signals produced a +115% cumulative return vs. +465% total return in SPX (including dividends).

Following both the Up and Down signals produced a -41% cumulative return vs. +465% total return in SPX (including dividends).

I focused on following the Up signals because following the Down signals was not particularly helpful. Down signals only generated a -72% ITD cumulative return vs. +465% SPX total return. Therefore, it isn’t really helpful to short the market when the signal crosses down. If you just sat out the market (aka only following Up signals and sat in 100% cash during the down signals), you only generated the +115% above. So we lost about 300% of upside by not participating in the market during the Down signals.

As we see there are plenty of signals per year:

I even tried other experiments such as:

• Only include NYSI signals that were < 0 (aka negative)
  • +22% total return, which is pretty terrible
• Only include NYSI signals that were in extremely oversold territory < -800
  • +9% total return
  • You might think, well, since these signals are rare, they must all be super accurate right? Nope, the false positive rate was around 50%.
• Sometimes the market is choppy and we get a NYSI signal that flip flops within a matter of days, or as few as 1 single day. I applied a filter of only including signals that were >=14 days (2 weeks) within each other for a more accurate reading.
  • +70% total return, which still lags behind putting on a SPX long on every single Up signal.

Additional Experiment Ideas / Notes

Please let me know if there was anything I missed, or if there are any other additional experiments that I could try.

• Does anyone have data on NYSI before June 1998?
• Buying SPXL (3x SPY) during Up signals, and returning back to 1x SPY during Down signals
• Selling 10% out-of-the-money covered calls during SPY Down signals.

Backtesting Process

My full notebook that I ran the backtesting on is here:

https://colab.research.google.com/drive/1rfNBs6zew0oEyIX2PbSornhMxMN1_otF?authuser=1#scrollTo=WX1U3pPuH2dY

1. Obtain raw price and price change data of SPY and NYSI since inception.
2. Created an EMA function that calculates the 8 EMA.
3. Create a ‘NYSIposition’ column that shows if its Below or Above the 8 EMA. Create a ‘Signal’ column that indicates whenever it Crosses from one side to the other.
4. Join the SPY and NYSI tables:

5. Make sure that all values tie with Stockcharts.com data.
6. Create a DaysBetweenSignals column that calculates the days between signals.
7. Calculate SPY_returns column that calculates SPY’s return based on each signal’s trigger return to the following signal.
8. Cumulative product the signals to get an annual return figure. Join it to an S&P 500 annual return table.
9. Produce alpha deltas and a cumulative return figure to see how the strategy did.