Using A.I. Bots to Pick Stocks

[Overkill]

Hello all,

I wanted to solicit the opinions of seasoned investors (Don and others) on this Forum about A.I. bots such as Scanz, Tickeron, TrendSpider, Equbot, and Google's "Bard" (which apparently gave incorrect answers to users in February). The use of A.I. bots is currently legal, but, though I am no expert, I wonder how investors' progressively increasing reliance on A.I bots will affect the performance of financial markets.

1) What are some advantages/disadvantages of using A.I. stock-picking bots?

2) Are they 'too good to be true?'

3) Would it be better to wait 4-6 months until more A.I. bot kinks get worked out?

4) Does this following approach sound feasible?:

a) Backtest 15 or so online A.I. bots for a few years each.
b) Then, 'forward test' on paper (with no financial risk) the 8 or so best ones.
c) Then, use 'old school' fundamentals (irrespective of A.I. bots), including consulting the opinions of professionals online, to pick possible stock candidates that are so-called 'inflation-proof' if the economy is currently experiencing an inflation. In the interest of diversification, perhaps pick 4 stocks from each Sector of the stock market.
d) Check each bot to see which stocks on my 'old school' 'short list' are on the greatest number of online A.I. bots' recommended lists.
e) When inflation is low, invest small amounts of money in 2 of my best, lowest risk 'short list' stocks from each Sector.

Thanks ahead of time!

[Norm]

I'm not a fan of back testing trading methodologies. There simply isn't enough data. Albeit, that's what funds always want to see. The entire book Technical Analysis from A to Z I always thought was a bit silly.

Now, an AI bot that has a possibility of working must include input from current news sources, a feel for economics, the public mood, current conspiracy theories, politics, listen to all the touts -- not to believe them, but to understand how those that do believe them might react -- understand that certain institutions heavily influence the markets. Of course, give them too much rein and they will influence the markets they use.

[beating vegas]

Hello all,

I wanted to solicit the opinions of seasoned investors (Don and others) on this Forum about A.I. bots such as Scanz, Tickeron, TrendSpider, Equbot, and Google's "Bard" (which apparently gave incorrect answers to users in February). The use of A.I. bots is currently legal, but, though I am no expert, I wonder how investors' progressively increasing reliance on A.I bots will affect the performance of financial markets.

1) What are some advantages/disadvantages of using A.I. stock-picking bots?

2) Are they 'too good to be true?'

3) Would it be better to wait 4-6 months until more A.I. bot kinks get worked out?

4) Does this following approach sound feasible?:

a) Backtest 15 or so online A.I. bots for a few years each.
b) Then, 'forward test' on paper (with no financial risk) the 8 or so best ones.
c) Then, use 'old school' fundamentals (irrespective of A.I. bots), including consulting the opinions of professionals online, to pick possible stock candidates that are so-called 'inflation-proof' if the economy is currently experiencing an inflation. In the interest of diversification, perhaps pick 4 stocks from each Sector of the stock market.
d) Check each bot to see which stocks on my 'old school' 'short list' are on the greates number of online A.I. bots' recommended lists.
e) When inflation is low, invest small amounts of money in 2 of my best, lowest risk 'short list' stocks from each Sector.

Thanks ahead of time!

I trade Futures not stocks. In Futures the CFTC has no rule against them. I don’t use them for trading reasons. It doesn’t suit my needs. I’m more about Capital Appreciation then Capital Preservation. From what I hear from people that do use them they have to keep adjusting them just as traders have to adjust their strategies.

You have to back test any trading strategies. Preferably with freeze frame replay.
If you decide use them make sure they adjust to a changing auction.
Currently inflation is very high. When inflation is high best time to not have money in a savings account because your currently losing around 10 percent a year on that money. I don’t even know why people have savings accounts no purpose in today’s economy. I just have a checking account.

Crude oil contracts have been a gift from gods lately. Just shorting the shit out of them.

But there are people who use bots or algorithmic trading who do very well.

[Overkill]

Thanks Norm and beating vegas. I had thought and hoped this thread would be much more busy.

Norm, regarding your perspective about back testing, do you believe that if a bot shows promise based on past performance, then its success on past performance was largely lucky due to, as you say, insufficient data?

beating vegas, forgive my ignorance as I am a novice, but by "a changing auction," do you mean a market constantly in flux due to constantly changing prices? (I just read more about the auction process, and I am thinking that's what you mean.). So, obviously, what I think you mean is to make sure any bot that I consult is open and learning on a daily (hourly?) basis and therefore is constantly updating its recommendations?

I plan to research what "freeze frame replay" means.

Finally, do you both believe that during our current high inflation is not a good time in general to begin investing in the stock market, or does it depend on the particular sectors or stocks? Perhaps during a high inflation is a good time to 'cut one's teeth' with regard to gaining experience with investing no money, just investing 'on paper.' Then, if one is consistently successful during a high inflation, perhaps they may tend to be even MORE successful in the future with a better economy?

[Norm]

Norm, regarding your perspective about back testing, do you believe that if a bot shows promise based on past performance, then its success on past performance was largely lucky due to, as you say, insufficient data?

Thing is, with any small set of data points you can come up with a winning strategy. Doesn't mean it will work in the future.

[beating vegas]

Thanks Norm and beating vegas. I had thought and hoped this thread would be much more busy.

Norm, regarding your perspective about back testing, do you believe that if a bot shows promise based on past performance, then its success on past performance was largely lucky due to, as you say, insufficient data?

beating vegas, forgive my ignorance as I am a novice, but by "a changing auction," do you mean a market constantly in flux due to constantly changing prices? (I just read more about the auction process, and I am thinking that's what you mean.). So, obviously, what I think you mean is to make sure any bot that I consult is open and learning on a daily (hourly?) basis and therefore is constantly updating its recommendations?

I plan to research what "freeze frame replay" means.

Finally, do you both believe that during our current high inflation is not a good time in general to begin investing in the stock market, or does it depend on the particular sectors or stocks? Perhaps during a high inflation is a good time to 'cut one's teeth' with regard to gaining experience with investing no money, just investing 'on paper.' Then, if one is consistently successful during a high inflation, perhaps they may tend to be even MORE successful in the future with a better economy?

First learn auction market theory,price action initial versus responsive. odds vs probability.I’m not a stock trader I trade commodities for full transparency. So the fundamental’s are going to be a little different because you are trading a business and need to understand what type of shape that company is in.

start with the basics. there are no shortcuts. also your a retail trader in the long term most retail traders lose money.

the best move for you learn the basics and instead of trading invest in some companies stocks that are in good shape make a nice spread of different stocks.

this is the advice I tell my relatives when they ask what to trade. it’s a full time job.

but good luck learn the basics study. there are no shortcuts.

[Archvaldor]

Hello all,

I wanted to solicit the opinions of seasoned investors (Don and others) on this Forum about A.I. bots such as Scanz, Tickeron, TrendSpider, Equbot, and Google's "Bard" (which apparently gave incorrect answers to users in February). The use of A.I. bots is currently legal, but, though I am no expert, I wonder how investors' progressively increasing reliance on A.I bots will affect the performance of financial markets.

1) What are some advantages/disadvantages of using A.I. stock-picking bots?

2) Are they 'too good to be true?'

3) Would it be better to wait 4-6 months until more A.I. bot kinks get worked out?

4) Does this following approach sound feasible?:

a) Backtest 15 or so online A.I. bots for a few years each.
b) Then, 'forward test' on paper (with no financial risk) the 8 or so best ones.
c) Then, use 'old school' fundamentals (irrespective of A.I. bots), including consulting the opinions of professionals online, to pick possible stock candidates that are so-called 'inflation-proof' if the economy is currently experiencing an inflation. In the interest of diversification, perhaps pick 4 stocks from each Sector of the stock market.
d) Check each bot to see which stocks on my 'old school' 'short list' are on the greatest number of online A.I. bots' recommended lists.
e) When inflation is low, invest small amounts of money in 2 of my best, lowest risk 'short list' stocks from each Sector.

Thanks ahead of time!

With respect you aren't using AI enough.

AI models are often aggressively stupid and just plain wrong. They give you no indication when they are wrong. Would you bet money on a human who offered advice like that? Watch less TV shows about how super-magic AI is and test it more.

With regard to the rest of your post you need to understand that any public information is factored into the market. Chat GPT is public info. Bard can barely count by the way. Maybe you could get an edge using cutting edge LLM models being developed on github or something before the market catches on but if tens of thousands of people are already doing something it won't work. You are using grenades to catch fish, works fine in theory till you blow your own fingers off.

TLDR: No.