Following the Signal Sigma Process

The approach to this article follows the step by step process described here. All visuals are sourced from various instruments available in the platform. If you are using the Portfolio Tracker, you’ll be able to see how we set it up for our own portfolio at the end of this article.

"When the music stops, in terms of liquidity, things will be complicated. But as long as the music is playing, you’ve got to get up and dance. We’re still dancing."

- Chuck Prince, Citigroup CEO, July 9, 2007

This is the spirit of the market nowadays. A pullback or correction in equities should not surprise anyone, yet bears are nowhere to be found. With risk markets (including the most speculative of asset classes - crypto) making new highs, why are we and our automatic strategies still in the market?

For one very simple reason: the music is still playing and we gotta keep dancing!

1. Asset Class Allocation

The first step in determining optimal portfolio positioning is taking a look at the performance of the main asset classes, and determining which are suitable for investment. The Asset Class Overview Instrument gives us a clear macro picture.

SPY has just completed a rare historical feat that’s usually a marker of recovering from a recession: a rise of +25% in the span of 5 months. This has only happened 10 times since the 1930’s.

While not an ominous signal in itself, the velocity of the rally is simply unsustainable. November to present, SPY has registered a 62% compound annual growth rate, which, if extrapolated to year-end, would take the benchmark ETF up to $750. That is obviously a science-fiction outcome with 0% probability of happening, so we need to ask ourselves how and why equities will eventually revert to a more attainable path.

The “Why?” is always a surprise. But a -5% to -10% correction in the coming months is probably the most realistic outcome, that would serve to work off the excesses accumulated during this run, and set SPY on a less steep CAGR slope.

Commodities (DBC) look to have established a base just below our stop level ($22.3) and are turning up. If this price action can be maintained by next week, DBC will become investible at these levels, with a rally target of $23.6.

(GLD) Gold’s rally has continued unabated in the past week, but it is getting very extended here, with a Sigma Score of 2.51. Our best guess would be that Gold takes a breather here, consolidates, and then rallies further.

The ratio of Gold to Gold Miners (GDX / GLD) tells us that the companies actually mining for gold have been lagging well behind the physical bullion itself. Over the last two years, gold has outperformed gold miners by about 25%. This is also the case in bitcoin (BTC-USD) relative to bitcoin miners. The takeaway is that costs to produce the commodity (labour, machinery, debt servicing etc.) have risen more than the commodity, leading to underwhelming profits for the producers.

It’s also illustrative to the fact that the two are distinct investments and it’s the reason why we separate stocks from commodities as asset classes.

TLT, the benchmark ETF for long dated government bonds, is now consolidating around support at $95 (S2) and so far it is selling off in reaction to CPI data. We’ll see how the situation develops further and whether a true breakout can occur.

Enterprise, our core investment strategy, will add to equity risk exposure today, following last week’s reduction. The adjustment is a not insignificant: +8.69% in SPY, bringing the total to 73%.

Bonds exposure is being reduced slightly to 22% today.

Some profit-taking is happening in Gold as well, with GLD getting sold to the tune of -1.63%.

Enterprise is also buying commodities, jumping the gun a little on the investability of DBC. The position is small, however, at 1.4%.

Cash goes from 8.6% to 1.3% today, with capital being almost fully deployed.

Since this model only trades 4 ETFs, we use it to judge overall portfolio positioning. The strategy is going a bit risk-on again, as technical signals continue to favor this approach. “If it ain’t broke, don’t fix it”.

2. Sector / Industry Selection

The next step in creating our portfolio positioning is to break down each broad asset class into more granular groups of assets. This will help us understand which pocket of the market is outperforming or underperforming and make our selection accordingly.

Since Equities are an investible asset class, we’ll take a look at how different Factors are performing and check for any notable opportunities.

We have included tables for this week and the prior 3 article editions in order to help you compare developments (click on the arrows or thumbnails to cycle through the tables).

The Dow Jones Industrial Average (DIA) has slipped to a negative Medium Term Trend, surprisingly - and given the current rally. All of the other factors are still trading up strong, in unbroken bullish advances.

In the short term, multiple factor ETFs look extended: Value Stocks (IVE), Mid-Caps (MDY), Momentum Factor (MTUM) and the Equally Weighted S&P 500 (RSP). Growth Stocks (IVW) are no longer overbought in the short term.

Among more granular Factor Returns, we’re noticing a pickup in Gross Profit Margin as a metric that’s favored by the market in the short term. Could this be telling us that inflation worries are starting to creep up again? This would make sense, as companies with higher gross margins usually enjoy a larger business moat.

On longer timeframes, the Pietroski F-Score ranks very well since it is a quality factor - and the market has rewarded quality.

Across all timeframes, the best performing factor is Net Debt / EBIT. Companies with high leverage have been absolutely on a tear lately. The market is telling us that if they are credit worthy, then they are doing something right.

Here’s how we stand on the Sectors front:

We have included 3 former tables from previous articles, for your convenience.

In the short term, Basic Materials (XLB), Industrials (XLI), Financials (XLF) and Utilities (XLU) - yes, Utilities - are overbought and extended. It’s very educative to look at how we determine these normalized deviations by studying the chart for XLU, where the 200-DMA is highlighted.

Note that the deviation above the 200-DMA is not that large in absolute terms. It’s just near the top of where this ETF is usually trading relative to it. In this case, the Sigma 200 score is measured at 0.79, or 79% of the highest percentage deviation recorded in the near term. Working with this transformation allows us to compare instruments across different asset classes and factors, where volatility is inherently different.

On a longer term horizon, most sectors are still relative under-performers to SPY. The exceptions to this are Real Estate (XLRE), Financials (XLF) and Communications (XLC). Healthcare (XLV) and Financials (XLF) could become prone to profit-taking soon, due to their unusually large deviations.

In another educational example, Tech (XLK) is no longer outperforming relative to SPY. How do we determine that? We take the relative price curve (XLK / SPY) and apply our usual regression technique. If the resulting curve crosses below the mean trend-line, it signals a loss in momentum where there previously was a gain. XLK has been able to recover from such a condition in the past, but we doubt that the sector will pull that off again.

These are early signs of a rotation from growth to value and from tech to… everything else non-tech.

Nostromo, our tactical allocation model, is only holding treasuries, via TLT, same as last week.

The model will also initiate a position in SPY on the next available BUY signal.

By holding on to 79% cash, Nostromo is taking on another risk - the risk of not performing in an up-market, as has been the case recently.

For more info about how Nostromo targets sectors or factors within a broader asset class, read this article. The first part sheds some light on the selection process going on in the background.

While underperforming in real life, this quirky model has its uses as a decision support tool. There is a clear case to be made here that equities are overbought. Nostromo is the only strategy to have almost zero drawdown during the Covid-19 crash in 2020.

3. Individual Stock Selection

Millennium Alpha, our star stock-picking algorithm has been on a tear lately, up 18.6% Year-to-Date and massively beating its benchmark (SPY). Drawdowns have been minuscule, with only -2.00% losses this year. Of course, this streak is too good to be true and will likely be broken, but it illustrates the power of a bull market coupled with a solid stock selection system.

Getting high performance from semiconductor stocks Broadcom Inc (AVGO) and NVIDIA (NVDA) is no surprise, but check out the gains realised on the following 2 stocks - Deckers Outdoor (DECK) and United Rentals (URI). Without the aid of a large data processing model, finding these stocks among other overhyped names would be very hard if not impossible.

As per usual, you can tweak this system using your own inputs if you wish.

4. Market Environment

The next step in our process is to take into account the type of market environment that we are currently trading in. For these purposes we use the Market Internals and the Market Fundamentals Instruments. Comments on the overall state of the market can usually be found in our Weekly Preview Article.

With markets reaching new all-time-highs, the divergence between SPY and the number of stocks trading above their 50-DMA and 200-DMA has become obvious. There was no recent activity to reverse this, so we’ll move this indicator to the “bearish” camp again.

Bearish Signal in Stocks trading above their 200-day Moving Averages

As a contrarian indicator, sentiment works best near extremes. Currently, there is enough skepticism in the market to fuel another run to new highs. Sentiment is not extreme yet.

Neutral Signal in Sentiment

The comparison of Z-Scores reveals the disparity between large cap performance (SPY) and the top 1000 stocks by dollar volume (the broad market).

Small and mid cap stocks are finally starting to show signs of life relative to their large cap peers. This performance difference has pushed the Z-Score differential lower in the past month, but there is a lot of work to do here in terms of “catching up”.

Bearish Signal in Market Internals Z-Score

Dollar transaction volume has surged well above the recent average during the last 2 weeks. While bullish for now, volume appears to be tapering off. If liquidity gets too low, the support behind this rally might cause “liquidity gaps” - sudden price movements and increased volatility. Both of which are not desirable.

Bullish (for now) Signal in Dollar Transaction Volume

5. Trading in the Sigma Portfolio (Live)

After reviewing all of the above factors, it’s time to decide on the actual investing strategy for our real-life portfolio.

We’ll continue this section by referencing the quote at the beginning: as long as the music is playing, this game of musical chairs (or charades, or whatever this is) must continue. And we need to stay in the market until “something breaks” and the trend changes. At that point, we will reduce exposure aggressively and even deploy hedges.

The second scenario that will get us to reduce exposure aggressively will be a further melt-up to $530 on SPY or getting an “Extreme Greed” reading on our sentiment indicator. For now, our automated systems are suggesting risk exposure is appropriate, and we don’t want to deviate too much from them.

Automated Strategies and Market Outlooks

The Sigma Portfolio (Live)

We’ll take the risk of slightly underperforming our benchmark in the next couple of weeks, as our exposure remains lighter than 60% stocks / 40% bonds. Our cash allocation sits at 14%, offering some cushion against volatility and dry powder to deploy in a downturn. We need to remain patient here and allow the market to dictate the next steps to us.

Click here to access our own tracker for the Sigma Portfolio and understand how the positions contribute to the overall exposure profile.

For now, there are no immediate trades that require our attention.

In total, we stand to gain $11.997 by risking $7.377 if our targets are correct. As the market goes higher, our risk / reward profile gets worse. Right now, the win-to-lose ratio stands at 1.62-1, lower than our general desired ratio of 2-1 (risking $7.000 should net us at least $14.000 in potential gains).

The factor exposure profile is balanced so as to shield us from potential profit taking in QQQ and MTUM. Instead, our profile leans more on Value Stocks (IVE) and the Equally Weighted S&P500 (RSP).

On the sectors side, exposure remains balanced as well, with intentional extra focus given to Financials (XLF).

If you have any questions, please contact us using your favorite channel. Have a great week everyone, and happy investing!

Andrei Sota