Kahneman-Tversky Portfolio Review: 31 December 2021

Terraced Agricultural Fields along the Rhine, Germany

The Kahneman-Tversky (K-T) Portfolio is a simple Dual Momentum portfolio that invests in only one of three possible assets available within the quiver. It does have an additional dimension of diversification in that two different lookback scenarios are used to measure momentum – a slow 252-day lookback for the Kahneman portion and a faster 60- and 100-day lookback combination for the Tversky portion.

Performance through 2021 looks like this:

with the total portfolio showing a small ~1% loss on the year. Obviously, this is very disappointing with the US equity markets showing a 29% gain over the same period. The K-T portfolio has the option of holding US equities through the VTI Fund, but was only invested in this ETF through the May-August period where we can see that performance was good (better than the benchmark AOR Equity-Bond fund). However, the use of Trailing Stop Loss Orders (TSLOs) took us out of VTI three times during the year resulting in us missing out on 3 months of bullish moves (black arrows in above figure) that cost ~3-4% per instance (~10% total) in terms of profits from a diversified fund (AOR) – and even more had we stayed in VTI.

Timing (Review Date) luck played a big role here since the TSLOs were triggered very shortly after the scheduled monthly reviews and re-entries were delayed until the next review date ~1 month later. In addition, the stop levels were obviously too tight triggering at ~4% level. Regular readers of this blog will know that I do not like using stop-loss orders – and this is an example of the reason why. However, I do require downside protection – that is provided (to some extent) by using the DM system – but I may have to accept larger drawdowns (up to ~20%) if I don’t want to miss out on strong bullish moves like those seen in the past year (but which we can’t expect to see again next year). I’ll discuss some ideas for hedging and exiting positions in a separate post dedicated to the subject.

This portfolio can be compared with Lowell’s McClintock, Pauling and Franklin DM portfolios – the first 2 of which are showing returns in excess of 20% and the later (with shorter momentum look-back periods) lower at ~6%. I don’t believe that Lowell has been using TSLOs with these portfolios. Obviously, luck can play a big role in determining the performance of these simple (relatively undiversified) portfolios.

Let’s take a look at rotations of the assets in the K-T portfolio using the “slow” Kahneman momentum lookbacks:

All ETFs lie in the bottom right quadrant of the (relative) rotation graph – i.e. have stronger long-term momentum than the reference TLT ETF (to the right of the vertical axis) but all also have weaker short-term momentum (below horizontal axis) – yet this short-term momentum is increasing (moving upwards). Since the Dual Momentum system focuses on relative strength (momentum) over the long term VTI would be the chosen asset to hold. This is confirmed in the Tranche recommendation sheet of the Kipling workbook:

Now we’ll move on to the rotation graphs using the faster moving Tversky lookbacks:

that moves the action towards the less desirable lower left quadrant with weak short- and intermediate-term momentum. VTI still shows positive intermediate term momentum and might be expected to generate a Buy recommendation, but:

checking the Tranche recommendation sheet we see that TLT is ranked number one (4th column from left). This is because volatility is included in the rankings and VTI is penalized for high volatility – but the choice is close.

Going into 2022 I will not be using TSLOs with this portfolio and I plan to check markets daily. I will not worry if I miss the check occasionally but, if I know that I will not be checking markets regularly, I may place a short-term TSLO (probably at the 1.65-2SD level). This also means that I will move from a regular monthly review to an adjustment schedule based on current model recommendations. For this simple DM portfolio, where changes are relatively infrequent, this should not be an onerous task. If adjustments are called for shortly after a review on this blog I will just report it in the comments section – otherwise I will write a review/performance update post ~monthly but not on any pre-planned schedule.

When the market opens for the New Year I will split the available funds and hold VTI in the Kahneman portion and TLT in the Tversky portion.

David