Post New

Back in July 2008 I first introduced the Mechanical Investing community to the NH-NL indicator that could be used to manage market risk. There are a number of ways to use this, but the method I originally presented subtracts the new lows from the new highs in the Nasdaq Composite daily, and then applies a nine day weighted moving average on the result. The original post is here.

http://www.datahelper.com/mi/search.phtml?nofool=y...

An alternative method is to use a thirteen day exponential moving average on the result. In this post I'll compare the two for the last few years using the QLD / QID as the investment vehicle. During BULL signals QLD is held, and during BEAR signals QID is held.

13-Day Exponential Moving Average

Date        Signal  Return   $100.00  % Gain
1/2/2018    BULL    24.69%   $124.69  24.69%
10/5/2018   BEAR    5.70%    $131.80  31.80%
1/31/2019   BULL    42.26%   $187.49  87.49%
2/27/2020   BEAR    -31.84%  $127.80  27.80%
4/29/2020   BULL    209.83%  $395.97  295.97%
11/22/2021  BEAR    47.38%   $583.58  483.58%
1/13/2023   BULL    0.00%    $583.58  483.58%
1/13/2023                             

9-Day Weighted Moving Average

Date        Signal  Return   $100.00  % Gain
1/2/2018    BULL    27.74%   $127.74  27.74%
10/4/2018   BEAR    8.45%    $138.54  38.54%
1/31/2019   BULL    58.08%   $219.00  119.00%
2/26/2020   BEAR    -24.99%  $164.28  64.28%
4/29/2020   BULL    213.61%  $515.19  415.19%
11/18/2021  BEAR    60.84%   $828.60  728.60%
1/9/2023    BULL    7.60%    $891.54  791.54%
1/13/2023                             

So if you are started with $100 invested on 1/2/2018 and put it into QLD you'd end up with $583.58 using the 13-Day EMA, while you'd do somewhat better using the 9-Day WMA resulting in a final amount of $891.54.

If you held QLD LTBH - no buys or sells from 1/2/2018 your gain would be 105.71% compared to 791.54% using the 9-Day WMA. In contrast, if you bought QID at the same time your loss would be -88.26%.

In other words, this does a bit better than just a risk management approach. A way to improve on this even more is to have a tool for identifying major market bottoms. Using such a tool you would not wait for it to return to a bull signal after going bearish, but instead, you'd enter based on the tool to identify a bottom. Jim has suggested signals from his tool over the years as one option for doing that.

Hi Zeelotes. Stockcharts uses NHNL data that gives similar signals using the 13-ema. There is sell on 2/4/19 (2 days difference) and the sell on 2/27/20 is the same. There are many signals in between using their data. I know you NHNL data is likely different, but how are you filtering signals that occur by just by using either the 9-wma or 13-ema? Thanks


13-ema from 1/1/2019 to 2/27/20

https://schrts.co/kMmNEmtW

Ironzi asked: There are many signals in between using their data. I know you NHNL data is likely different, but how are you filtering signals that occur by just by using either the 9-wma or 13-ema?

I make full use of two things to limit whipsaw:

1. Four year cycle research - you can read about it here from 2/13/2006: http://www.datahelper.com/mi/search.phtml?nofool=y...
2. Consensus of Long-Term Indicators - all my many posts back in the day detail some of these indicators. I combine them to determine times when it requires going defensive at the first sign of trouble ahead, and times when the consensus is super bullish and signals like these from NH-NL are ignored.

Using these methods my actual signal dates and returns using the 9-Day WMA are:

Date        Signal  Return   $100.00    % Gain
1/2/2018    BULL    27.74%   $127.74    27.74%
10/4/2018   BEAR    52.20%   $194.43    94.43%
12/24/2018  BULL    114.00%  $416.07    316.07%
2/26/2020   BEAR    38.62%   $576.74    476.74%
3/16/2020   BULL    399.41%  $2,880.30  2780.30%
11/18/2021  BEAR    79.57%   $5,172.02  5072.02%
10/11/2022  BULL    10.22%   $5,700.67  5600.67%
1/13/2023                               

I've uploaded two spreadsheets to my blog that illustrate how to calculate the various methods I've outlined above. I've also included links to three distinct sources for the data. Of course, you'd have to obtain the data daily and add it to the spreadsheets to continue to track these indicators.

I also show one of the tools I use to mark a transition point at bottoms - from bearish to bullish.

Zee

I am unsure of the meaning of your response to Ironzi. Do the table you first posted have whipsaw reduction? If yes, is that why the WMA performs better than the EMA?

13-Day Exponential Moving Average

Date Signal Return $100.00 % Gain
1/2/2018 BULL 24.69% $124.69 24.69%
10/5/2018 BEAR 5.70% $131.80 31.80%
1/31/2019 BULL 42.26% $187.49 87.49%
2/27/2020 BEAR -31.84% $127.80 27.80%
4/29/2020 BULL 209.83% $395.97 295.97%
11/22/2021 BEAR 47.38% $583.58 483.58%
1/13/2023 BULL 0.00% $583.58 483.58%
1/13/2023


9-Day Weighted Moving Average

Date Signal Return $100.00 % Gain
1/2/2018 BULL 27.74% $127.74 27.74%
10/4/2018 BEAR 8.45% $138.54 38.54%
1/31/2019 BULL 58.08% $219.00 119.00%
2/26/2020 BEAR -24.99% $164.28 64.28%
4/29/2020 BULL 213.61% $515.19 415.19%
11/18/2021 BEAR 60.84% $828.60 728.60%
1/9/2023 BULL 7.60% $891.54 791.54%
1/13/2023

Thanks

Craig

Craig asked: I am unsure of the meaning of your response to Ironzi. Do the table you first posted have whipsaw reduction? If yes, is that why the WMA performs better than the EMA?

No, the two original posts are the same in this respect.

Thanks Zeelotes,

I might need some coffees but I have tried the wma 9 ndnh-ndnl but I get much more signals than what you do so the tables of your first post of the thread are with your other indicators whipsaws reduction? Right?

Post New

Mechanical Investing FAQ

---

Contact Shrewd'm
Contact the developer of these message boards.