How to make strategy private?
Topic Starter
Fri Jan 4, 2019 - 07:10

Sorry if this is written somewhere but I can't seem to find a way to not have my strategy public. I want to see how it goes before and if I ever make it public. I can't see an option in any of the settings.

How do I keep it unpublished with the new site please?

Fri Jan 4, 2019 - 08:22

Sorry but WealthSignals does not have "private sandboxes" for a long time.
Fri Jan 4, 2019 - 09:13
I will discuss with the team if we should bring back private strategies or not. Thank you.
Sat Jan 5, 2019 - 05:42
I'm all for adding back private sandboxes.
Topic Starter
Sat Jan 5, 2019 - 21:25
Thank you for your responses. There are I'm sure many reasons why a strategy might want to be private but here are mine:

1) The strategy setup is not overtly difficult but it is not intuitive and there is a learning curve. The last thing I want is for me to be messing around learning publicly. If I end up creating and deleting a strategy several times it looks bad and reduces credibility should I ever make it public.

2) I want to get used to posting daily if I do this before I go live.

3) I can't imaging anyone buying into a strategy with 0 out of sample results so this gives some incubation time where the data is being updated daily and the results are actually out of sample but they aren't displayed publicly until they've been proven. Again, it's a credibility thing. I wouldn't want to share a strategy with bad out of sample results, especially if there are potentially multiple strategies.

4) While going through the learning curve, I'd rather not be flagged as a bad author publicly. I created a strategy on the site but haven't done anything with it yet and am already getting bad boy emails. I'm not flagged yet but am assuming I will be soon.

5) The programming for it could potentially be very easy: add a Public checkbox on the settings page, add a single column to the database, Public (Y/N or 1/0), keep everything else the same including the out of sample results, and simply condition the listing on the strategies page to those with a Y/1.

So if you could add that Public option please that would be great.

Thanks for considering it.
Sun Aug 9, 2020 - 23:50
I was baffled to discover that there's no option to make a strategy private. I second this request for all reasons above, specially #1, #2 and #4.

I fail to see any downside of providing an option to make a strategy private (specially for strategies with no subscribers).

On the same lines, a "Closed for new investors" option is a must.
Mon Aug 10, 2020 - 04:00
If strategies had to have backtests before playing when I got involved, I motion not to allow strategies "play time" unless either a backtest is submitted or all strategies are to remain public so that screenshots do not mislead when they do become public.
Fri Aug 21, 2020 - 01:38
I'm not sure if I understand your comment.

Strategies are tracked out-of-sample, and this is reflected in the return chart. Having a backtest (and being public or private, for that matter) doesn't make any difference for a investor.

Again: I fail to see any downside of providing an option to make a strategy private (specially for strategies with no subscribers).
Sun Oct 4, 2020 - 07:57
Again: I fail to see any downside of providing an option to make a strategy private (specially for strategies with no subscribers).

If it's not allowed to be made public later, isn't it? To avoid potential cherry picking.
Thu Jan 21, 2021 - 03:16
What's the downside of author cherry picking?

I'm strategy author, and if start with, say, 5 private strategies to test their performance out-of-sample, and find out that only 3 make sense for whatever reasons, then decide to publicize only those 3 strategies, then why bother?

Today you can do that anyway -- just set you subscription rate to the maximum allowed so that they do not attract subscribers, and after some time cherry pick the ones you actually want to market by changing the price to a reasonable value, and cancelling the others. There is has a huge downside for WS IMHO in that it pollutes the strategy list with loads of expensive strategies with short/bad OOS stats.
Thu Jan 21, 2021 - 20:49
Someone could publish the best performing strategies from private strategy pool and say "Hey, I have *real* trade data to prove they work great." However, they are best in hindsight. Their *real* trades are actually no better than backtest data but misleading in this case.
Thu Jan 21, 2021 - 23:07
> Someone could publish the best performing strategies from private strategy pool and say "Hey, I have *real* trade data to prove they work great."

That would be great for WealthSignals, investors, and the author, wouldn't it?

Anyway, I'm not sure if understand your comment about hindsight. From an investor evaluation POV, a profitable strategy running for some time on OOS data with a reasonable # of trades is *nowhere near the same* as a backtest over the same period. First and foremost, someone can just cheat on a backtest ("if (DAY == '2020-03-21') then BUY"), something they could *never* do on OOS real-time data. Also, trading in real-time is more challenging in many cases -- for example, an author could have a strategy that opens 10 positions/day+close on next day, which looks wonderful on backtest, but wouldn't be feasible to trade simply because they don't have the power/time/stamina to enter/close that amount of orders every single day, year in, year out.

Also, AFAIK WS's objective is not to evaluate traders, but strategies. IMHO investors are (should be) interested in strategies with a good OOS track record, and that's the author deliverable.

Lastly, I can just paper trade those 10 strategies in Excel for 3 years and, after that period, add only the 3 most successful ones to WealthSignals. That's the same exact thing, except I won't have a *proven* track record on OOS data, confirmed by a third-party (WS). I would have to wait 3 more years for that. Who loses: WealthSignals, investors, authors. Who wins: nobody.
Sun Jan 24, 2021 - 16:26
A story I heard is: a monkey can throw darts on the wall to choose which stock to invest. In this example, in all the different outcomes of throwing darts, he must be able to win great on one or two which were proven by *past* *real* market. Now, he presents the two great picks having been proved right to the public, and claims that he has the skill of picking the right stocks which has been proved in the *real* market, are you going to believe what he claimed basing on the two great picks he disclosed? This is the misleading or cherry-picking that were talked about in previous comments.

In summary, out-of-sample history of a cherry-picked strategy should not be more trust-worthy than back-testing history.
Mon Jan 25, 2021 - 10:36
Agree with MarketStudentYG. That's the same as the stock picking newsletter scams. You send out a different series of picks to different groups, and subsquently follow up with the group that got the winners.

The backtest, however, isn't supposed to convince you to trust a system. Instead, what we hoped to accomplish was to show customers that the out-of-sample WealthSignals test is a "continuation" and "rhymes" with the in-sample test. You can see this on a lot of successful strategies like "Deterministic Knife Juggler", "Saltamontes", and "ETF Pairs Arb." (This would be more obvious if we had an option to show the equity curve in Log scale.) When you see that the WS results don't "rhyme", then probably the system was over-optimized.
Tue Jan 26, 2021 - 04:12
Let's agree to disagree. Again: to me that's no different from an author paper trading using Excel, and publishing only winners strategies. In fact, to a certain extent we all do this in that we only publish strategies that pass backtest/paper-trading testing.

Anyway, I rest my case. Will leave for WS admins to decide. There are plenty of other live tools for paper-trading that one can use to certify OOS trades, and at the end of the day is up to investors to decide. I just think it would be perfect for everybody that it could be done in WS for the sake of transparency, but we all now that the world is not perfect.
To reply to this topic, you need to login first.