Your strategy sounds ideal for somebody like me who has also recently retired early. The equity curve on your chart is remarkable not least for the fact that you have predicted all 3 of the last bear markets and got out of equities almost as soon as they have begun and then continued to make good profits through each of them. Can you tell me what it was about a quite normal January 2020 that made the system think that it was a good idea to go fully defensive? It does sound almost too good to be true that you have swerved all of these bear markets and made good profits at the same time. In 2022 there were not many places to make money apart from general commodities but your system got through that period very well.
Thanks Tim. It’s a very fair question, and I’d make an important distinction: the system didn’t “predict” COVID or any of the other bear markets. Each month-end decision is made systematically using only information that was available at the time. There’s no forecasting of crashes involved.
COVID is actually a good example. At the end of December 2019 the portfolio still held Nasdaq and UK property alongside gold. At the end of January 2020 it held Nasdaq and the S&P 500 alongside long-duration US Treasuries. So it certainly wasn’t sitting there in January somehow knowing that a pandemic and market crash were coming. As conditions changed, the portfolio continued to adapt, with healthcare subsequently replacing one of the equity positions while Treasuries remained on the defensive side, and gold later returning.
2022 is interesting for a different reason. Commodities featured heavily on the defensive side for much of the year, while the equity selections also changed as relative strength shifted. Later in the year gold and cash became increasingly important as fewer risk assets qualified.
That’s really the key to how I think about the system. It isn’t trying to identify the next bear market at all. It responds systematically to what is actually happening across a deliberately diverse set of assets. There will inevitably be occasions when it becomes more defensive and subsequently turns out not to have needed to and that’s part of the price of trying to control large drawdowns.
I’m deliberately guarded about the precise inputs and thresholds because that is the proprietary part of the strategy, but I’ll be covering more of the robustness testing in future posts, including out-of-sample behaviour, parameter sensitivity and the dangers of overfitting. I think those are entirely reasonable things to scrutinise when a backtest looks unusually strong, thanks.
Your strategy sounds ideal for somebody like me who has also recently retired early. The equity curve on your chart is remarkable not least for the fact that you have predicted all 3 of the last bear markets and got out of equities almost as soon as they have begun and then continued to make good profits through each of them. Can you tell me what it was about a quite normal January 2020 that made the system think that it was a good idea to go fully defensive? It does sound almost too good to be true that you have swerved all of these bear markets and made good profits at the same time. In 2022 there were not many places to make money apart from general commodities but your system got through that period very well.
Thanks Tim. It’s a very fair question, and I’d make an important distinction: the system didn’t “predict” COVID or any of the other bear markets. Each month-end decision is made systematically using only information that was available at the time. There’s no forecasting of crashes involved.
COVID is actually a good example. At the end of December 2019 the portfolio still held Nasdaq and UK property alongside gold. At the end of January 2020 it held Nasdaq and the S&P 500 alongside long-duration US Treasuries. So it certainly wasn’t sitting there in January somehow knowing that a pandemic and market crash were coming. As conditions changed, the portfolio continued to adapt, with healthcare subsequently replacing one of the equity positions while Treasuries remained on the defensive side, and gold later returning.
2022 is interesting for a different reason. Commodities featured heavily on the defensive side for much of the year, while the equity selections also changed as relative strength shifted. Later in the year gold and cash became increasingly important as fewer risk assets qualified.
That’s really the key to how I think about the system. It isn’t trying to identify the next bear market at all. It responds systematically to what is actually happening across a deliberately diverse set of assets. There will inevitably be occasions when it becomes more defensive and subsequently turns out not to have needed to and that’s part of the price of trying to control large drawdowns.
I’m deliberately guarded about the precise inputs and thresholds because that is the proprietary part of the strategy, but I’ll be covering more of the robustness testing in future posts, including out-of-sample behaviour, parameter sensitivity and the dangers of overfitting. I think those are entirely reasonable things to scrutinise when a backtest looks unusually strong, thanks.