Determinant portfolio: situation as of 01.12.2023

What's new with the allocation strategy

After a series of backtests, I managed to optimize the overall functioning of the portfolio by slightly modifying the adaptive allocation criteria. I found that the overall key performance indicators became better by:

• focusing on absolute momentum (relative momentum now no longer has any impact on positioning and disappears from the summary table)
• fixedly and strategically weighting assets, rather than varying them according to their volatility (the latter also disappearing from the table)
• removing the QQQ ETF (see reasons below)

My backtests therefore partly call into question the results obtained for the adaptive allocation, as presented in my book. It must be said that, in its original version, the latter includes more asset classes and does not include cryptos. In addition, it is not concomitant with other investment strategies, with which there may be overlaps. The approach used here is therefore significantly different from the initial allocation. It is therefore normal to adapt its operating criteria.

To avoid confusion, I will henceforth refer to this strategy as “Tactical Asset Allocation” or “TAA”, rather than “Adaptive Allocation”.

Monthly performance in CHF

In detail, here is the performance of the PF, of each of its strategies and of its benchmark index, the MSCI Switzerland, during this last month:

The portfolio has achieved a very good overall result, even if it is below the market. It should be noted that the latter has shown a very strong rebound over such a short period. The portfolio's performance is all the more remarkable, given that the cash share fluctuated between 21 and 39%. In addition, it was heavily penalized by the dollar, which depreciated by 3% against the CHF, dragging down values denominated in greenbacks. It should be remembered that short-term movements in foreign currency assets tend to fade over the long term.

There QVM strategy persists in negative territory. The underweighting initiated in recent months has limited the damage to the entire portfolio. Since the beginning of this year, QVM stocks have been largely losing out compared to other approaches. Does this mean that this strategy has had its day? Nothing could be less certain. The method has been tested over several decades and, even if it has always experienced many difficult moments, it has always recovered strongly, obtaining over the long term a profitability much higher than the market as a whole. 

There Blue Chips Strategy had a magnificent month of November, even managing to do better than the Swiss market. Taking into account the weakness of the dollar against the franc, this is really a very satisfactory result.

Same for the Trading Auto Signal. This result confirms the adaptive nature of this approach, which knows how to hold its own in bullish, bearish or trendless markets. This is supported by backtests and results obtained in the past.

Finally, the TAA strategy performed well, nothing more. Cryptos performed very well, but this increase was wiped out by the decline in gold. Moreover, this approach was unfortunately taken against the market, by closing QQQ at the beginning of the month, misled by its momentum (which was then fading). The Nasdaq nevertheless rebounded powerfully just after. This sometimes happens when using trend-following indicators like moving averages.

About Moving Averages

Most of the time, this trend-following indicator works well to detect market reversals. However, sometimes, moving averages induce false signals. This is the price to pay to avoid major bear markets. Research and my own backtests also highlight that their use does not do much better in terms of long-term profitability than a simple buy-and-hold strategy. On the other hand, it allows to significantly reduce volatility, which allows to obtain a better Sharpe ratio (return relative to risks).

Other changes within the portfolio

That being said, I've been thinking about QQQ for quite some time now. Since the reintroduction of Trading Auto Signal, this ETF has been a duplicate, as it is so correlated to SPY (coefficient 0.84), due to a very similar composition (QQQ on the left, SPY on the right below).

The signal is much more accurate than QQQ in tactical asset allocation, because it reacts on a daily basis rather than at the monthly close. In addition, it is based on a multi-criteria assessment, and not on a simple moving average. As a result, I have decided to remove the line dedicated to the Nasdaq in tactical asset allocation. The Trading Auto Signal will gradually be reinforced as compensation.

For those who absolutely want to continue investing in QQQ by following the monthly signals, they can continue to do so based on the QVM strategy. Indeed, this line uses the S&P 500 to define its positioning. Since this index is highly correlated to the Nasdaq 100, the signals generated by the moving averages are practically identical. Moreover, when the trend in the US unemployment rate is favorable, the two approaches are perfectly equivalent.

Finally, due to the strengthening of the Signal vis-à-vis the tactical asset allocation, the total cash position of the portfolio is now mainly derived from the short sale of SPY or the non-investment of the Trading Auto Signal. Since these movements are mainly short-term, it is no longer wise to use intermediate-term treasury bills (ETF IEF) to invest part of the excess cash. This line is therefore now also removed. On the other hand, under liquidity, the currencies will now be specified (USD or CHF), depending on the short-term momentum of these currencies.

Complex simplification

For several months, the portfolio has been undergoing a process of "complex simplification", aimed at focusing on the most relevant assets and techniques, both from a profitability and risk perspective. It also aims to make the portfolio coherent and intelligible as a whole. This simplification is taking place without ever deviating from the principles derived from research and my own backtests. That is why I call it "complex". The number of asset classes has decreased, some indicators have been removed, which has allowed (and will continue to allow) to limit the number of transactions, without inducing an increase in risk (quite the contrary). The face of the portfolio thus appears simpler than before, but its brain and muscles remain highly efficient.

Year-to-date performance in CHF

In detail, here are the results of the PF and its reference index, the MSCI Switzerland, during this year:

The performance of the PF, slightly lower than the MSCI Switzerland (but better than the Swiss Performance Index at 2.5%), is nothing extraordinary after 11 months. It must be said that the strength of the franc against the dollar (+5.2% since January 1) has weighed throughout this year. However, helped by its lower volatility, the portfolio still maintains a better Sharpe ratio than the market. With equal risk, the determining portfolio therefore performs better than the market.

Determinant portfolio and MSCI Switzerland index since the beginning of the year (performance in CHF)

Performance since 2010 in CHF

Since its launch in 2010, The portfolio maintains a slight lead over the Swiss market.

MSCI Switzerland portfolio and index since 2010 (performance in CHF)

Portfolio Update

The determining portfolio will be updated very soon, in the members section, with the changes mentioned above.