Market Timing Models – Backtesting Breadth Signals and Focusing on the Big Swings

Today we will dive into breadth indicators and test a modified version of the Index Breadth Model here at TrendInvestorPro. First, however, I will review the S&P 500 SPDR as it toys with its 200-day SMA here at month end. In particular, I am monitoring upswings in four key major index ETFs. After the breadth dissertation, I will update the High-Low Lines and Bullish Percent Indexes, which also hold short-term keys, and finish with the yield spread and Fed balance sheet.

Monitoring the Current Upswing

I talked about a make or break week last week and the S&P 500 followed through with another gain this week. This move puts the index above its 40-week and 200-day SMAs. It is also the last trading day of the month and a monthly close above the 200-day SMA could trigger some trend following systems.

Does this sound the “all clear”? The market is never clear and the evidence now is mixed at best. SPY is up some 35% since the March low and needs another 12% to forge a new high. Despite this historic advance, the total number of new highs is anemic and the percentage above the 200-day EMA is not indicative of a bull market.  

At this point, I think we are dealing with a market of swings and need to focus on the current upswing. The March crash was the big downswing and the subsequent rebound is the big upswing. SPY consolidated the first two weeks of May and then broke out of this consolidation. This breakout keeps the swing alive and puts the 200-day SMA in play.

The current upswing is bullish with three levels to watch going forward. A close below 285 would fill the 18-May gap and negate the flag breakout (strike one). A close below 280 would break the flag low and forge the first lower low (strike two). And finally, an RSI move below 50 would mean the momentum cup is half empty (strike three). The chart below shows upswing support for QQQ, MDY and IWM.

Even though the Index Breadth Model is bearish, the Sector Breadth Model is slightly bullish, the trend since late March is up, key yield spreads continue to narrow and the Fed continues to expand its balance sheet. The Index Breadth Model has yet to trigger bullish and this keeps me neutral at best. The upswing since March holds the key and I will be monitoring this upswing closely.

Backtesting a Breadth Model

I tested a version of the Index Breadth Model with a 16+ year backtest and will share preliminary results today. StockCharts data only goes back to 2010, which we can only test signals during the bull run from March 2009 until February 2020.

The indicators in the backtest are similar, but a little different because I am using breadth indicators from Norgate data. Instead of %Above 200-day EMA (StockCharts), Norgate calculates %Above 200-day SMA. Instead of 52-week highs/lows based on intraday highs/lows (StockCharts), Norgate bases 52-week highs/lows on closing prices. For reference, StockCharts notes 19 intraday highs in the S&P 500 on Thursday and Norgate notes only 12 closing highs.

I am using Norgate data with Amibroker to run the backtest. Also note that the data are based on historical constituents. This means breadth data from March 2008 reflects the constituents  that were in the index at that time. This is very important because the S&P 500 looked a lot different in 2008 than it does now. Click here to see changes over the years.

First, let’s set a basic benchmark by looking at returns for buy-and-hold and a 5/200 day SMA cross. Buy-and-hold returned 8% per year, but with a 55% drawdown in 2008-2009. Buying and selling based on the 5-day SMA crossing the 200-day SMA returned 6% per annum, but with a much lower drawdown (21%).

The chart below shows adjusted SPY with signals using the moving average crossover (red/green arrows). The histogram shows when the 5-day crosses above/below the 200-day (green/red).

The next table shows returns using the Index Breadth Model signals based on Norgate data. The Compound Annual Return (CAR) is 7% and the Maximum Drawdown (MDD) is 23.4%, which occurred in 2010 because of whipsaws. The May 2010 flash-crash brought on this volatility and there were eight swings of 7% or more from April to September.

The chart below shows the S&P 500 with signals over the last 16 plus years. I am showing the S&P 500 on this chart so we can see the “true” price levels in the past. Keep in mind that dividend adjustments literally change past prices and historical values do not reflect the true values at the time. The Index Breadth Model signals range from +9 (all bullish) to -9 (all bearish). The model signals bullish when it is net positive (at least 5 of 9 indicators with bullish signals). The model triggers bearish (red) when at least 5 of 9 indicators have active bearish signals.

One signal jumps out here. The Index Breadth Model turned bearish on 30-July-2007 and the S&P 500 hit a new high in October 2007. The index then traded flat until yearend and broke down in January. Needless to say, the period from August to November would have been challenging because the model was on a bearish signal as the S&P 500 hit new highs. A non signal to consider is sharp rebound after the bearish signal in August 2015.

I could say that we are in a situation like 2007 or 2015, but these are only two data points and these are not enough to base a solid conclusion. Currently, six indicators are bearish (-6) and three are bullish (+3). This means the Norgate Index Breadth Model stands at -3 and still bearish. As with any model or strategy, there are going to be good signals, bad signals and whipsaws.

Norgate versus StockCharts

There are no new signals on the Index Breadth Model. The six long-term indicators remain with bearish signals, while the three breadth thrust indicators remain with bullish signals. The breadth thrusts in April triggered with strong upside participation, but follow through has been limited because High-Low Percent is barely positive and the percentage of stocks above the 200-day EMA has yet to clear 40%. Leadership is still relatively scarce (new highs) and there are still a lot of downtrends out there.

On the lag issues, StockCharts triggered bullish on 5-February-2019 and Norgate did not trigger bullish until April 2nd, two months later. Norgate triggered bearish on 5-August-2019 and StockCharts followed suit on 15-August. The rest of the signals are largely inline and I imagine that these differences are negligible when looking back 15 years or more.

StockCharts Index Breadth Model Update

Despite another gain in May, there is no change in the Index Breadth Model because we did not get follow through signals from the High-Low Percent and %Above 200-day EMA indicators.

The breadth thrust indicators for the S&P MidCap 400 and S&P SmallCap 600 moved back above +30% this week, but the breadth thrust indicator for the S&P 500 fell short. This is the second non-confirmation from the S&P 500. The small and mid cap breadth thrust indicators have been above +30% three times since April 9th, while the S&P 500 breadth thrust indicator only made it above once.  

The High-Low Percent indicators remain anemic as not one cleared the 5% level this week.

The %Above 200-day EMA indicators continue to improve with the S&P 500 nearing 50%, the S&P MidCap 400 hitting 40% this week and small-caps exceeding 30%. Despite these improvements, the cup is still half empty for the S&P 500.

Click here for an article and video explaining the indicators, signals and methodology used in the Index Breadth Model. This article also includes the signals of the last five years.

High-Low Lines Still Rising

Despite relatively few new highs, there are even fewer new lows and this is pushing the High-Low Lines higher. The chart below shows the S&P 500 and S&P MidCap 400 High-Low Lines above their 10-day EMAs since mid April and the S&P SmallCap 600 High-Low Line turning up this week. These indicators are short-term bullish as long as two of three are rising.

S&P 500 BPI Back above 60%

The S&P 500 Bullish Percent Index ($BPSPX) moved back above 60% and the big three are in bull mode. The S&P 100 Bullish Percent Index ($BPOPX) and Nasdaq 100 Bullish Percent Index ($BPNDX) did not break below 40% in mid May and kept the group bullish.

Sector Breadth Model is Unchanged

There is no change in the Sector Breadth Model. Nine of the eleven breadth thrust indicators are on bearish signals, while eight of eleven High-Low Percent and %Above 200-day EMA indicators are on bearish signals. The bullish breadth thrusts occurred in April, as did five of the six other bullish signals in High-Low Percent and %Above 200-day EMA. The only notable signal in May was Tech %Above 200-day EMA (!GT200XLK) moving back above 60%. Outside of this signal, there were not any bullish additions this month.

The Sector Breadth Model is technically mixed as the three biggest sectors battle the fourth, fifth and sixth biggest sectors. It is a 3-sector tag team battle for the ages. Will tech, healthcare and communication services prevail or will consumer discretionary, industrials and finance drag the bigger sectors down? I am betting on the latter at this point.

Yield Spreads and Fed Balance Sheet

Investment grade bond spreads narrowed further over the last few weeks and the AAA spread is back near the pre-crisis highs (blue lines). AAA bonds are the best of the best (blue chips). BBB bonds, which represent the lowest tier of investment grade bonds, are still above pre-crisis levels, but moving in the right direction (down and narrowing).

Junk bond spreads continue to narrow and this is positive for stocks because it shows continued improvement in the credit markets. The Junk bond spread is still well above pre-crisis levels, but moving in the right direction (down and narrowing). This means the spread between junk bonds (risk assets) and Treasury bonds (safe-haven) is tightening. CCC spreads are also moving lower, but to a lesser degree.

The Fed balance sheet expanded a mere $60 billion this week and this is the smallest expansion since mid March. The green arrow line shows total assets creeping higher from September to February and then surging from March onwards. While I cannot verify this, I would guess that some of this money is finding its way into riskier assets (stocks).

Thanks for tuning in and happy Friday!

-Arthur Hill, CMT
Choose a Strategy, Develop a Plan and Follow a Process

Scroll to Top