Weekly Wrap-up: The Power of the 4-Year Cycle
The following article was originally published at The Agile Trader on March 26, 2007.
The Dynamic Trading System's auto-traders took gains of +7% and +9% respectively on E-mini NDX and SPX futures contracts last week. The System's position gains total +378% since the portfolio's launch on July 15, 2005 (net of all auto-trade commissions and subscription, clearing, regulatory, and Exchange Fees). The System's Model Portfolio has netted +115% in total return during this period.
We have had some questions regarding the nature of the difference between Net Position Gains and Total Return. Position Gains measures the sum of percentage gains/losses on the margin required for each trade.. Total Return tells you how much money the account has gained in total relative to its starting point.
You can see all the specifics at Agile Trader Index Futures Service: Performance.
Many (if not most) trading services will only discuss Position Gains because those numbers can get to be absurdly high. But they won't discuss Total Return because that involves describing an asset allocation model and being responsible for both the concept and application of money management. (As we discussed at some length last week, money management is probably the first, second, and third most important thing in trading.)
For instance, suppose I recommend trading a portfolio of $100,000 in 10 different equally weighted positions. Now further suppose that of those 10 positions half of them gain +10% while the other half lose -5%. Not bad, right? In terms of position gains we will have (5*10%)+(5*-5%)= +25%.
However the gross total profit will be (5*$1,000)+(5*-$500)= $5,000-$2,500 = +$2,500. Which is a +2.5% total return (before we net out the cost of 20 transactions, which at, say $15/trade comes to $300, leaving us with a net of $2,200, or +2.2%).
Hmmmmm. +2.2% isn't nothing, but it's a far cry from the +25% figure you'll see in the promotional email you receive. And if you get a promotional email from a service that put on sets of such positions they may claim to have made +125% while your net total return is more like 11.5% even with compounding (1.022^5-1 = 11.5%).
Now you see why we think money management is so important in trading? I hope so. And as important as it is ALL the time, it may be especially more so over the coming 6 or 7 months. And here's why.
This chart shows us the SPX's percentage gain off its October '02 low (red). It also shows the SPX's median "4-Yr Cycle" performance over the past 44 years (blue). We are currently in the 11 th 4-yr cycle since 1962. Each 252-trading-day period on the chart below represents roughly a year.
Now, there are a few things that stand out on this chart:
- The red line and the blue line enjoyed a better-than +0.93 correlation (extremely positive) through trading-day # 817 ( 1/9/06 in this cycle) and since then the correlation has been -0.12 (slightly negative).
- The median performance tends to top out around trading-day 800 in the cycle (we're now at trading day 869).
- The Standard Deviation from the median (dark grey squiggly lines) tends to remain fairly narrow through about trading-day 645 and then begins to widen out, exploding outward from about day 825.
- The sharpest gains tend to be made in the first 115 days of the cycle.
- The sharpest losses tend to come in the last 200 days of the cycle (in which period we are now).
- Median performance for roughly the first 28 months is +47% while median performance for roughly the last 20 months of is slightly negative with a median draw-down of about -12%.
OK, median-shmedian. How about those exploding Standard Deviation lines, what kind of distribution do they represent?
This chart tracks each 4-year cycle, beginning with whatever was roughly the October low of year-one in the cycle. (Start dates range from 10/1 to 11/19 but 9 of the 11 lows came in October.)
- First thing I see is the 2 cycles that enjoyed strength until way over on the right side of the charts: 1986 (brown) and 1998 (powder blue). Impressive, right? Anyone remember the crashes of October 1987 (delayed by a year but a WHOPPER) and 1998 (right on schedule)? Makes those rallies look a little scary.
- Next I see the 4 cases in which the market rolls over just after day 504 (2 years in) and cascades lower until, in each case, all (or almost all) of the prior 2 years' gains were given back.
- Then I see the 5 iterations of what we'll call "middling" scenarios. No particular blow-off to the upside, but no marked crashes by this point in the cycle (this point in the cycle being represented by the orange vertical line). The SPX appears to be playing out one of these middling scenarios in the current cycle.
- Finally, what are the fates of the other middling cycles? Each one declines to AT BEST +49% by day 911, which in this cycle will come right around Memorial Day, 2006.
- Beyond that point, 2 of the prior middling cycles rallied before galumphing down into the next October low and 2 of the prior middling cycles continued on down, cannonballing to the bottom of the pool.
In summary, the first 2 years of the cycle show extremely consistent positive market behavior with every cycle but one enjoying at LEAST +30% in appreciation through the first 500 days. (The exception was 1987, which followed the 2 nd strongest cycle (brown) and collapsed a year late.) Meanwhile the last 2 years of the cycle show extremely spotty performance during which there have been a couple of unstable, dangerous upside blowoffs, but which have otherwise displayed either flat or negative performance.
It sure looks as though, if an upside breakout were going to occur before the fall of '06 it would have happened already. And it also looks like if there IS a significant breakout, it will very likely be part of a scary and unstable bout of manic euphoria that will ultimately lead to an ensuing bout of depressive dyspepsia for the bulls.
So, let's suppose that the 4-Yr Cycle is going to exert its pull on the SPX and will, between now and October, take the index down to a price no better than +49% above its low of October '02. Here's what that would look like:
Not an "uncompelling" first target, 1157, as it represents a flirtation with the highs of late '01, early '02, and late '04, as well as the lows of April '05.
As far as Fibonacci retracements go, 1161 (not shown above, but very close to our 1157 target) is the mid-point between the SPX high of 1553 in March 2000 and the bear market low of 769 in October 2000, so that area on the chart has a lot of attractive magnetism.
If 1157 were to break to the downside, Standard Fibonacci retracements of the 2002-2006 rally (38, 50, and 62%) yield further downside targets from current levels to 1100, 1040, and 980 respectively.
Am I calling for the next secular bear market? NO, I AM NOT! Indeed, in the absence of an upside blowoff between now and October, the cycle data suggest that the SPX is likely to appreciate in excess of +30% between like an October '06 low and October '08. But in the natural "yin and yang" process of breathing in and out, of creating and then destroying to make room for further creation, the SPX is getting toward "overdue" for a creatively destructive event. And if that event doesn't come at roughly its appointed time, then the delaying of it itself is likely to exacerbate the extremity of it.
Now, before we leave the subject of the potential for a negative event, I'd like to show you another chart that is surprisingly compelling.
Here we plot the SPX's Forward Price/Earnings Ratio (blue) against its annualized rate of appreciation for the subsequent 2 ½ years (30 months). The correlation is -0.88, which is about as good an inverse correlation as you'll find in nature...which is to say that over the past 20+ years the market has netted a very positive response 30 months subsequent to dips in the PE and it has netted much less positive or negative responses 30 months subsequent to spikes in the PE.
Now look at the spike on the blue line (PE) in January '04. And 30 months subsequent to that date? July '06, which is the beginning of our target period for a cyclical low on the SPX. So, if the correlations on this chart continue to hold up, we'll see the SPX dip toward its Jan '04 high near 1140 before rallying at least toward the point in time that's 30 months subsequent to Dec. '05, which is mid 2008).
Could either or both of these paradigms fall apart? You bet. None of our work is "deterministic" in any philosophical or scientific sense. It's all probabilistic. And while we can always be wrong, it appears that the high-odds scenario is bearish over the coming 6-7 months and then bullish for about 30 months after that.
**** **** ****
Slowing earnings growth is the most obvious issue for the stock market.
SPX EPS grew +23.7% in CY04 and about +13% in CY05. The current consensus calls for EPS growth of +10.8% in CY06 and for +5.2% in CY07.
Forward 52-Week EPS estimates are decelerating in a range that is likely to become problematic soon.
The F52W EPS consensus now calls for +13.5% growth, down from 20%+ a year ago. The 3-month annualized rate of growth in the consensus is +6.4% and is likely to lead the Y/Y figure lower. As the yellow highlights indicate, when the blue line is either declining at a level below +10% or is below 0%, the market tends to struggle. This combined with the still-flattening yield curve...
...should make it difficult for the market's PE to expand.
The spread between the Effective Fed Funds Rate (4.64% and soon to be 4.75%) and the 10-Yr Treasury Yield (4.675%) is down to 3 basis points. And this coming week the 10-Yr Yield will rise or the curve will invert, neither of which will be perceived as bullish for the stock market.
I can't get my mind off this next chart:
The correlation here is shockingly high. We looked at it last week but we should keep this in our minds. When the Y/Y Change in the Fed Funds Rate declines faster than Y/Y EPS growth, the Fed is being friendly. When the FF Rate is holding stead in the face of relatively speedily declining F52W EPS Estimates the Fed is being unfriendly. Currently we have an unfriendly Fed. And the notion that they will take stop raising rates before something big in the economy breaks is probably naïve. (Or as one commentator put it last week, they probably won't stop until someone is carried out on his shield.)
The Fed's Fair Value calculations divides the F52W EPS consensus by the 10-Yr Treasury Yield: $86.40 /0.04675 = 1848. But we can all probably agree that that's not very helpful
Our Risk Adjusted Fair Value is calculated like this:
F52W EPS / (TNX + Med ERP)
F52W EPS = Forward 52-Week EPS ($86.40)
TNX = 10-Yr Treasury Yield (4.675%)
ERP = SPX Earnings Yield – TNX (6.63% - 4.675% = 1.96%)
Med ERP = Median Post-9/11 ERP (1.96%)
$86.40 / (0.04675+ 0.0196) = 1303
The SPX is now fully valued relative to current long-term interest rates and the median level of fear in the stock market since 9/11.
Long-term interest rates are either going higher or going below the Fed Funds Rate. As of Tuesday there will be no space between the Rock and the Hard Place. As the Yield Curve goes upside down, it will be interesting to see whether the market can maintain or even increase its joie de vivre or if it will succumb to darker impulses.
**** **** ****
A MODICUM OF BULLISH TA
On a more jovial note, the market has been displaying some internal strength in the face of all the dark winds I've been elaborating upon.
Both the NYSE Advance/Decline Line and Cumulative Volume Line are confirming the SPX's break above 1300 by setting new highs of their own. So, while we continue to see cause for mid-term concern, we do not at this point have the kind of negative divergences on these 2 important internal indicators to tell us that the market has to roll over immediately.
We'll be keeping an eye on these indicators (as well as on the New High Percent indicator that we've been looking at daily in our Morning Call). And once this trio of internal measures rolls over we'll be looking for the bearish cyclical scenario to play out.
If you would like a free 30-day trial to our daily market analysis, please join us at The Agile Trader. And if you are interested in our application of the Dynamic Trading System to the Futures markets, you can read more about the System at The Agile Trader Index Futures System.
Best regards and good trading!