In our globalized complex world, "uncorrelated asset classes" do not exist. That's your answer. Hope this helps you further: In devising a risk management strategy... first, figure out what would ruin you...then eliminate/mitigate. -->
Economic theory literature has let people down many times. Look @ gold during 2008 financial crisis. It "wasn't supposed to do that". Be suspicious of everything. Don't listen to academics. Put simply, no one really knows anything. The only thing valuable of knowing is knowing that you don't know. With that being said, act accordingly.
From a Quant's standpoint? Correlation and Non-Correlation is one of the FEW tools that you should pay attention to. If I was under the gun, and told that I could only pick three ways to measure a market? Hrm. 1) Standard Deviation. 2) Correlation / Non-Correlation. 3) Price Strength of a Move. And honestly, both of my Trading Interests ... Firms, are basically built on the above concepts. And I'm not unique or special by any means. Many other firms do this too. That's it. With those three, I can pretty much do whatever I want. If you are looking at Correlation and Non-Correlation? Pretty simple really. Just define your periodicity, and measure it. Almost any platform out there, has that tool for you. ANY asset class will move to tightly correlated (approaching 1), and away from correlation (contextually, whatever "loose" is between the two assets). Some asset classes like Stocks and Bonds are INVERSELY correlated by their nature. To your original post, yes, that would essentially be one trade, say within a monthly periodicity (20 days). To help work against getting your trades too correlated, and without getting too ... what one of my Partners calls: "Quanty" ? LOL Check your innate biases, about market direction. Why does it have to be one or the other? Ok, you want a risk-off bet. Great. Ok, let's say you're long Gold. Why not have a Long Equities, risk-on position on as well? Well, because I think the market is going to go down. Ok, no problem. But do you have a net-net positive expectancy process, that looks in the other direction? Because if you do? You don't have to care what happens in the markets. Take a guy that has three processes he runs. Let's say he runs a trend process. A Mean Reversion Process, and a Pairs Trade Process. His trend process is long Equities, when it's time to be long Equities. Ok. Then, let's say he runs a Currency Mean Reversion Process. Great. Then he runs a pairs trade process, and he get's long Gold, because it's calling for a mean reversion. He's still long equities. But he's also long Gold, for a mean reversion play. He has no idea which is going to win, nor does he really care. He's running both a risk on, and a risk off bet. And he knows all of his processes are net-net profitable. But he definitely does not want to be Long Gold, Long Siver, Long Bonds, Long Swissy ... because yeah, he then essentially has one trade on.
Hard to disagree with any of the above. Do you find it hard to find markets with correlation close to 0? Basically something in the -10 to +10 range.
Crypto and small cap stocks. Vape stocks for example are currently responding to regulatory changes not the broader market.
Financial markets and the horses ! more broadly, financial markets and sporting markets, and sporting markets amongst themselves (intra-sporting markets)