[CoachKandSportsguy]

This type of machine learning (ml) results will eliminate most investment analysts jobs, and reduce a portfolio manager's job to strategy automation.

Note the portfolio analysis results over 30+ years.

the most important results are:
length of backtest
investment instruments (etfs good, which etfs?)
annualized return (not return from inception)
Sharpe ratio: greater than 1.0, which is return is greater than risk taken
Max drawdown: the worst loss from the maximum gains when the portfolio ml is wrong.
Worst year: when the model is wrong during the year, what is the annual return look like? does it adjust rapidly to the conditions?

If and when anyone interviews an investment manager with a program,
ask for the results listed above, if you can't get them, you really are only getting a sales pitch for a retail client.

The screen shot is the picture of the Twitter post, click on the url if the screen shot doesn't load, its a small screen shot so not sure why it doesn't show.
https://x.com/WifeyAlpha/status/1824034705689194726

[Stu from NYC]

There is value in sitting down with the management of company to be invested in.

Also can a computer understand the future of a line of business?

Good tools but not the entire picture

[CoachKandSportsguy]

Originally Posted by Stu from NYC

There is value in sitting down with the management of company to be invested in.

Also can a computer understand the future of a line of business?

Good tools but not the entire picture

There was a paper, I'm not sure I downloaded it, but put all companies financial ratios and results through historical analysis with ML, and with enough history, can get a list of investable companies and non investable companies. Then one can visit management, but management is always biased. Text based ml analysis of company messaging between competitors has been used for quite awhile.

And the portfolio here is made up of ETFS, index replication with actual index stock components, so individual company analysis is not going to make a difference. Old style portfolio management of buying and selling individual companies has fallen to the wayside with custom traded ETF portfolios.

you want 3 x leverage on the SPX index long, SPXL
you want 3 x leverage on the SPX index short, SPXS
you want the top 50 largest SPX companies quality, XLG, goes up faster than SPX
you want invest in the utility sector with diversification, XLU
the list goes on and on

So that is why the investment industry is moving away from individual stocks, and moving towards portfolio of etfs for tax advantaged diversification, acting on industry analysis without individual stock purchases, and getting in or out easily and quickly. .

[CoachKandSportsguy]

Now, with T+1 settlement, getting in an out of a position for reduced rebalancing holding periods, means ML trading can be more automated, and less waiting on fast down days when you get out and then want to get right back in again the next day, or even the same day. . and with trading hours from 7 AM to 8 PM every day, except options, . . . there is plenty of time to get one's positions bought or sold. .

[New_user]

In addition to that we a small investors have no way of knowing if we invested in the next Enron or if the books are signed off on by a fraudulent accounting firm (Arthur Anderson).

Originally Posted by Stu from NYC

There is value in sitting down with the management of company to be invested in.

Also can a computer understand the future of a line of business?

Good tools but not the entire picture

[Stu from NYC]

Originally Posted by CoachKandSportsguy

There was a paper, I'm not sure I downloaded it, but put all companies financial ratios and results through historical analysis with ML, and with enough history, can get a list of investable companies and non investable companies. Then one can visit management, but management is always biased. Text based ml analysis of company messaging between competitors has been used for quite awhile.

And the portfolio here is made up of ETFS, index replication with actual index stock components, so individual company analysis is not going to make a difference. Old style portfolio management of buying and selling individual companies has fallen to the wayside with custom traded ETF portfolios.

you want 3 x leverage on the SPX index long, SPXL
you want 3 x leverage on the SPX index short, SPXS
you want the top 50 largest SPX companies quality, XLG, goes up faster than SPX
you want invest in the utility sector with diversification, XLU
the list goes on and on

So that is why the investment industry is moving away from individual stocks, and moving towards portfolio of etfs for tax advantaged diversification, acting on industry analysis without individual stock purchases, and getting in or out easily and quickly. .

Back in my days as a credit analyst, thought there was value in sitting down with owner and/or controller over how the company is doing as well as its prospects.

Very often my gut feeling about management was proven correct.

[CoachKandSportsguy]

Originally Posted by Stu from NYC

Back in my days as a credit analyst, thought there was value in sitting down with owner and/or controller over how the company is doing as well as its prospects.

Very often my gut feeling about management was proven correct.

yes, totally agree with that for the job, but talking about investment portfolio management of billions of dollars. . you can't scale to that level and be profitable. . . cost to manage portfolio is too high

But with huge amounts of money, one person can now manage billions with ETFS and Machine learning algorithms.

wsj.com

behind paywall, but just the words visible gives enough information . .

[dewilson58]

Give me five minutes and I'll make up a BS historical return graph and give it a cute name.

[dewilson58]

Oh, here is a review:

John
1 review
GB
Rated 1 out of 5 stars
Feb 22, 2024
Completely fraudulent company with terrible financial advice
This company is run by a "financial expert" on Twitter and advises people on how to trade and macro market conditions. Performance has been terrible and if you question anything you are blocked and your comment is removed from their Twitter feed.

Other clients are also complaining but any questions or slightly negative comments get blocked and questions are never answered, all this for a significant quarterly fee. The stats they post on Twitter are cherry picked to show good performance, the reality is the performance is poor and often very wrong, costing their subscribers a lot of money in the process.

It is a completely disgraceful way to run a macro and financial service company and I wish I had never come across this person. They are also constantly saying they do everything for their followers but it is a technique designed to lure in potential new fee paying subscribers. Please be very careful before listening to or paying for any advice from this person as it will cost you a significant amount of money in bad financial advice.

Date of experience: February 12, 2024

[LoisR]

Does that mean that every time your AI "manager" decides Its time to sell, or buy, a new stock you'll incur a capitol gains sale to impact your taxes? Same as for a financial manager, not worth the expense.

[Stu from NYC]

Originally Posted by LoisR

Does that mean that every time your AI "manager" decides Its time to sell, or buy, a new stock you'll incur a capitol gains sale to impact your taxes? Same as for a financial manager, not worth the expense.

Yes or if they take a loss, but that reduces cap gain.

[rsmurano]

a couple of things:
you talk that etfs are the golden goose, they aren't as a whole. You still have to know what you are doing. Some etfs are great, some are dogs. IMO, the biggest advantages of etfs are you can trade them like stocks, and they should be a little cheaper.
All that history you talk about is built into most stock screeners today, going back to at least 10 years, not sure why you would want to go back more than that.
Your AI etfs of the future are still actively managed which costs you more taxes, higher expenses, and you are at the mercy of a machine which is right now the guy who programmed that machine. You still have initial programmers who program AI code so the machines will act smarter over time, which you have some of that today. For example, when the market hits a certain drop level of the day, you have machines set to sell (or possibly buy) and that's why you have a run on the market. You can (and you should) setup trailing loss for your accounts so you take the emotion out of selling and you won't end up with a stock/etf down 40%.

I don't like anything about a "managed fund", I don't care if it's the so-called best manager or some machine making the decisions. I always use index funds which will beat any managed fund over the long run. Managed funds have a hurdle from start: you have to pay many people to look into which companies to buy initially, then you have these same guys looking when to sell, which all of this costs a lot of money. Then most of these managed funds have a high turnover (I've seen over 400%) which costs you money in transaction costs plus short term gains expenses on your taxes. On the other hand, Index funds have very low low cost (some of mine are .02%), very low turnover rate, and its over a broad base of companies.
The comment: "Old style portfolio management of buying and selling individual companies has fallen to the wayside". I beg to differ. I made a fortune during the past year or so getting into stocks like Tesla, Apple, Nvidia, chipotle, and others that went up hundreds of a %. I have done this with apple probably a dozen of times in the past decade. I wait until a stock goes out of favor for some reason (you know these stocks will not go under) and you pounce, then when you feel that the company is at its peak in value, you sell. So what if it continues going up, in my mind, pigs get slaughtered. I bought apple in the $120's then sold it a while back in the $180's. It went sideways for a while then started going back up. I'll just wait until another bump in the road or another stock split and jump back in. I have Meta around $95-$100 18 months ago, now take a look. I got out when I ws up in the low 300's of a %. Now its up over 400%.

I'm going to give AI related programs/machines a few years to get perfected before I will trust them with my money. I wouldn't even use chatgpt right now. Apple's "Apple Intelligence" looks good but even then I won't trust it right now either.

[DAVES]

Every trade there is a buyer thinking it is a good time to buy at a price and a seller who thinks it is a good time to sell at a price.
Information is supplied by people who say you need us.
We think we are investing not gambling
At the track there is or used to be THE RACING FORUM that you would purchase.
The market? we are not gambling-REMINDER, .There are many RACING FORUMS for sale.
Typically I compare my returns to the S&P many funds, etfs etc do.
An awakening. SPY is Apple 6.9%, Microsoft 6.7%, Nvidia 6.2%, Amazon3.69%,Meta class a 2.24%
Alphabet class a 2.17%, Hathaway class b 1.71%, Broadcom 1.51%, Tessla 1.39%

[rsmurano]

My s&p index fund has gone up the same amount this past year. Since you mentioned technology holdings inside of spy, look at some of the technology index funds. So far this year my tech index fund is up over 31% with an expense of .09% and pays a dividend. It might be a little late getting into these now

[mntlblok]

Originally Posted by CoachKandSportsguy

yes, totally agree with that for the job, but talking about investment portfolio management of billions of dollars. . you can't scale to that level and be profitable. . . cost to manage portfolio is too high

But with huge amounts of money, one person can now manage billions with ETFS and Machine learning algorithms.

wsj.com

behind paywall, but just the words visible gives enough information . .

I'm so ashamed, but SIL turned me on to this "archive" thingy that often bypasses paywalls - especially if the article has been out for a few days. . .

What Does Nevada’s $35 Billion Fund Manager Do All Day? Nothing - WSJ