Hello Everyone,
Does anyone have an opinion on why there was never a designer model for MAG7? My analysis shows picking the top 2 MAG 7 at any time over the last 15 years outperforms the market by a lot. I'm pretty sure it would be one of the top designer models almost every year except 2022. It would also be a great allocation to a book. The other great thing about the MAG7 is they are very liquid and you can use options to manage the trade or sell premium.
Cheers,
Mark Vincent
I believe the term is only a year old? And because it groups stocks that soared (that's how they got to have such massive market caps), it is biased. Any strategy that starts with a group of past winners isn't probably not something people are going to invest in no matter how good the backtest.
That's just my thoughts on why no designer models on it yet.
I looked at this at the end of 2021.
My SA article "Higher Ranked is not Better" shows that investing at the end of the year in the stocks that made the highest contribution to the performance of the S&P500 over the calendar year is not a good strategy and would only approximately match the performance of SPY.
https://seekingalpha.com/article/4477445-higher-ranked-is-not-better
Agree on the part that such a strategy is highly biased and is doomed to fail out of sample.
Disagree on the part that people would not invest in it though. Sadly enough.
I think it's a completely different way of looking at building a model. Every single one of those stocks has beat the S&P over the last 5 years. So throwing a dart at the board you win. It's really the same bet as just putting your money in the Nasdaq. We all know the Nasdaq has beat most designer models over the long run. My analysis shows that just using momentum with the top 2 Mag 7 beats the Nasdaq. There is no fundamental ranking system used. Has anyone else tried this and if so what were the results?
I do agree it is a group of stocks and knowing when to drop one of them and add another would not be an easy task. That being said I'm sure you could come up with some rules on when to add and when to subtract. It's not pure mechanical investing but it's better then adding and subtracting with no rules.
I think the whole issue here is that you would not have known what the 'MAG7' stocks would have been at the moments in time that you backtested on. We only know this looking back now.
To get a reliable system, you would have to create a universe with rules that determine what stocks make up the group of stocks you chose between. You could pick the largest stocks, perhaps the largest technology oriented stocks (though you would have to really ask yourself whether you would not be using knowledge from the present moment there as well). But MAG7 would in a sense be fooling yourself in my opinion.
Yes it really is a belief in the largest most dominate tech companies. If you think 2 out of the 7 during some rebalance period in the future are going to underperform the market then yes it's a bad idea. But if you had to make a bet I think it would be a good one. Time will tell. It would almost be the same as saying I bet any new designer model will outperform the Nasdaq over the next 10 years. I would not take the bet.
