Andrew Ang: Should AI Agents Run Your Asset Allocation? | #653

So maybe let's talk about diversification all the way down.

The old adage is that diversification is the only free lunch, and it's true.

It is really the only free lunch.

And you want diversification at all levels in your portfolio as much as you can do.

We've applied that concept to factor investing in three levels.

So the first layer of diversification is using multiple signals to measure a given factor.

So you don't just measure value with book to market or earnings to price, right? We wanna measure it with intangible capital and economic value added and have a proper multi-dimension view of these factors.

And you can do that for quality, momentum and other factors too.

The second layer of diversification is that you don't want to just have one factor in your portfolio.

Value is a great example because it went through a horrific, absolutely horrific drawdown from 2018 to 2020, really bounced back at the end of 2020, going into 2021.

But it was the worst value drawdown in history and over a century of growth.

of the return financial experience.

So let's hold other factors as well.

When you want to hold them in a balanced way, I put them with low tracking errors that can be a core part of your portfolio.

Now, the factors do vary over time, and you, I think, might think about factor tilting or positioning for these factors.

Sometimes that's a little bit controversial, but I think you can do it.

But you don't want to rely on just one single method to position your factors or vary them over time.

So we're using three models, a market similarity model, which places more weight on episodes in the past that are similar to today.

You can look at factor momentum or how well these factors are doing themselves.

These factors have trends.

So yes, there is a momentum of momentum.

And then finally, you can look at their dispersions or how attractive the economic opportunity is in these factors.

And you can measure that by the dispersion of these different factor measures.

Simply put, if there's a big difference between the returns of the stocks going up versus stocks going down, I think momentum is more attractive at that time.

If there's a large difference in valuations between the expensive stocks and the cheap stocks, that also means it's a more attractive time for value.