OTHER INVESTMENT MODELS --> | SuperMODEL | The Big Mac | Cash/Equity Yield Model |

The Long-Bond Yield Curve (LBYC) Investment Timing Model is the eighth timing model developed by PowerStocks Labs for the South African investment community. Like our flagship SUPERModel Composite Market Timing System, it is also a
long-term investing timing signal. The timing service is available to ENHANCED/PRO subscribers only.

Currently our only long-term investment timing model is the SUPERModel. Over a 32 year period this strategy traded 9 times, giving an average of one trade every 3.5 years. However this strategy was only vested in the JSE 46% of the time and its trades averaged 19.6 months duration with the longest trade 37 months (the last 2003-2007 bull run) and the shortest only 4 months. Although SUPERModel dodged just about every major bear market in the last 32 years, most trade exits occurred before the 3 year minimum holding period for shares deemed by the tax-man for profits to qualify as capital gains (normally 10%) as opposed to personal gains at the marginal rate (40%).

Whilst the incredulous SUPERModel gains more than made up for this "tax penalty", there are fund managers and investors that want the middle ground between a pure buy-and-hold strategy and the SUPERModel. Something that still gives the ALSH a walloping but stays invested in the JSE for the longest time possible to minimise portfolio disturbance and its associated tax implications and only heads for the exits when really big trouble is looming. A timing model which is the "last one to switch off the lights when it leaves the building." Hence the LBYC was borne to fill that gap.

The LBYC timing model uses the relationship between the average yield on a basket of long-term (10-30 year) government-backed bonds (such as the R157 and R186) and the average yield on a basket of short-term debt instruments such as 3-month treasury T-bills, to offer very infrequent and profitable buy/sell investment signals on the JSE. This relationship is known as the "Yield Curve".

The LBYC strategy has just over one-half the trade rate of the SUPERModel, with only 5 trades in the last 26 years versus the SUPERModels' 9 trades over the same period. Every trade exceeded the tax threshold of 36 months and averaged 55 months (4.6 years) with minimum 42 months (3.5 years) and a maximum of 84 months (7 years). The strategy posted a whopping 12,008% return since 30 June 1983, or 20.3% CAGR by being invested in the JSE 72.56% of the time. During the same period, the JSE ALSH index returned 2,658% or 13.6% CAGR.

Note that at PowerStocks, when calculating returns on timing strategies we include buy and sell brokerage costs of 0.75%, we exclude taxes and we exclude dividend payouts in the return calculation. When the timing models are "out the market" we deem hibernation in fixed interest investments (call account or money market funds) earning prime less 4%.

For your interest, the SUPERModel returned 23,532% or 23.4% CAGR during the same period but obviously with 9 trades versus 5 and the returns of all but one trade subject to tax at the investors' marginal tax rate. Another thing to bear in mind is that the SUPERModel was only vested in the JSE for 47% of the time versus the LBYC at 72% which would imply the LBYC should also accumulate more in dividends. So although it appears the LBYC strategy delivered 50% of the returns of the SUPERModel, when once takes these considerations into account it is probably more like 70%.

We will now delve into the theory behind the yield curve. We implore you to wade through the next section, even if you struggle a bit. We have tried to keep it as simple and easy to understand as possible. It is vital for you and your future financial wealth to understand the next section. When you understand the next section and couple it with the tools we are going to provide you as part of LBYC, you will have promoted yourself to the top 20% of "informed" private investors on the JSE.

A bond is a debt investment in which an investor loans money to an entity (corporate or governmental) that borrows the funds for a defined period of time at a fixed interest rate (coupon) to finance a variety of projects and activities. The indebted entity issues a bond that states the interest rate (coupon) that will be paid and when the loaned funds (bond principal) are to be returned (maturity date). Bonds are commonly referred to as fixed-income securities and are one of the three main asset classes, along with stocks and cash equivalents. Government backed bonds are considered the lowest risk investments next to cash (the probability of the government reneging is low.)

The yield of a bond is simply the coupon rate (interest rate) divided by the market price of the bond. If the bond price rises, the yield decreases and vice-versa. Essentially, the relationship between yields on long-term government-backed bonds and short-term government-backed debt instruments such as treasury bills is what is known as the "Yield Curve". Typically, short-term yields (or interest rates) are lower than long-term rates, so the yield curve slopes upwards from left to right, reflecting higher yields for longer-term investments. This is referred to as a "Normal Yield Curve".

When the yield curve is positive, then long-term bonds offer higher yields than shorter-term treasuries since investors are ploughing money into the stock markets in anticipation of positive outcomes on the economy. Longer term bonds thus have to offer a premium to attract investors in them, and because of the laws of supply and demand, long-term bonds will be decreasing in price. This means their yields will be increasing.

When the spread between long and short term yields rapidly expands, this is known as a "Steep Yield Curve" and is quite a bullish signal for the stock markets. However when the spread between short-term and long-term interest rates narrows, the yield curve begins to flatten. A flat yield curve is often seen during the transition from a normal yield curve to an inverted one and also signals a period when investors are undecided about the future prospects of the economy and interest rates.

When the relationship is inverted however, this is known as an "negative yield curve". It means there is a high demand for longer term bonds, driving their prices up and thus lowering their yields. Furthermore, to finance their purchases of long-term bonds, investors will liquidate their holdings of short-term instruments and their yields will in turn rise. The shifting of funds from short-term to long-term securities in anticipation of a recession will cause the yield curve to invert. In this case, short-term interest rates exceed long-term rates and the flight to the safety of bonds is due to investor concerns about the future of the economy which results in money being yanked out the stock market resulting in JSE selling pressure and lowered ALSH returns.

INVERTED YIELD CURVE : An interest rate environment in which long-term debt instruments have a lower yield than short-term debt instruments of the same or similar credit quality. This type of yield curve is the rarest of the three main curve types and is considered to be a predictor of economic recession. Sometimes referred to as a "Negative Yield Curve."

Historically, the appearance of inverted yield curves have predicted many of the U.S. recessions. In the 1990s, Duke University professor Campbell Harvey found that inverted yield curves have preceded the last five of six U.S. recessions since 1970. Due to this historical correlation, the yield curve is often seen as an accurate forecast of the turning points of the business cycle and hence of major stock market corrections.

In general then, for the investor one could summarise the use of the Yield Curve as :

1. If the yield curve is inverted, the investor should not buy long-term assets and stay liquid.
2. If the yield curve is flat he/she is indifferent, but would probably be best off staying liquid
3. If the yield curve is normal he/she is indifferent but would probably be best off staying invested.
4. If the yield curve is steep he should definitely be invested.

There, that wasn't so bad was it? In the next section we examine the behaviour of the yield curve in the South African context and reveal one of the most powerful forecasting tools devised by us yet.

We take nothing for granted, so we set out to test the Yield Curve theory in the South African landscape together with its relationship with the JSE. To do this we assembled a basket of long-term government-backed debt instruments (10-30 year bonds) and a basket of short term instruments (such as 3-month treasury bills) and tracked their average yields versus each other on a monthly basis since June 1986. Unfortunately our database for these various instruments never stretched further back than this, but the period is sufficiently long and includes 5 major corrections and 4 business cycles to make it statistically relevant. The result appears below in the monthly time-line:

One can immediately observe a visual correlation between secular (long term) JSE bull markets and a positive yield curve relationship (where the red line is above the green line). It also appears visually apparent that besides the crash of October 1987 (which is regarded as a "Black Swan" event and had nothing to do with the economy or recessions) the 5 periods of inverted yield curves corresponded to periods of lacklustre (when it would be better to have your money in fixed-interest instruments) to downright awful JSE returns.

When we compared the periods the yield curve became inverted, to previous SA economic cycle down-swings, we discovered that the curve became inverted prior to ALL 4 previous down-swings and economic recessions although the timing between the inversion and the recession dates has varied from two to four quarters. Since the JSE is itself a leading indicator of economic cycle turning points we are more interested in the relationship between the inversions and the JSE, which as we can see from the chart are well correlated.

We also observe that the relationships between the average yields of the two different baskets of debt are relatively stable and do not "whipsaw" you easily into or out of the stock market. If one includes the 1987 crash, we can say that this model predicted 4 out of 5 (80%) crashes and 4 out of 4 (100%) "recessionary" periods, pretty close to the results in the Duke University research.

While all of this suggests that the slope of the yield curve is a valuable leading economic indicator and correctly predicts future economic conditions, its real use for investors is providing early short term warning of JSE corrections (which are themselves predicting the economic downturns.)

You will note that the yield curve inversion itself (where the red and yellow lines actually cross over each other) do not exactly mark the LBYC "BUY" and "SELL" signals demarcated by the left and right edges of the transparent blue shaded boxes. This is because our analyses has revealed a certain "tolerance" in negative or positive yield curve spreads and only after this threshold has been exceeded do things "give in". For example a negative yield curve of 0.5% does not really push things over a cliff. Similarly a positive yield differential of 0.25% doesn't really jolt the JSE into a bull market.

The threshold we observed is a consistent number and forms a nice 26-year "neutral line" and is obviously proprietary to our timing model (we won't be telling anyone what it is!) Just using the simple yield-curve crossovers would deliver a strategy that produced 9,900% return, but by incorporation of our "neutral threshold", the returns improve 21% to over 12,000%. In addition to the improved performance, the threshold relaxes crossovers which would have introduced some whipsaws in the 1990-1992, 1995-1996 and 1998 time periods when the two lines were tracking each other very closely. Thus spurious trades, exactly what we are trying to avoid with this timing model, are eliminated. (Otherwise we could have just as well gone to the SUPERModel and doubled our returns!)

Further investigation into the existence of this  "yield spread tolerance" produced a valid explanation - some economic purists define a yield curve inversion when the spread between long-term yield and short term yields is below the market reward for interest rate risk at the time. This makes perfect sense and will explain why only after our threshold is exceeded, do the markets make a move.

So for this reason, when we speak of yield curve inversion its when the red line falls below the yellow line by more than the market reward for interest rate risk. Additionally when we plot yield spreads (defined as the difference between the red and the yellow line) we centre this plot around a neutral line defined as zero subtract market reward for interest rate risk. When the spread falls below this line we have a "true yield curve inversion" which serves as the LBYC SELL signal. Similarly when the spread exceeds the neutral line we have a "true positive" yield curve that serves as the LBYC BUY signal.

And that is why the edges of the blue shaded boxes representing periods we are vested in the JSE are slightly offset from the cross-overs of the red and yellow lines in the previous chart, and why sometimes we have the red line below the yellow line but no SELL signal (such as from late 1996 to early 1998.)


Some other casual observations we can make about the LBYC timing signals:

1.It only bailed half-way through the 2002 crash, but it bailed during a period when the market was going sideways and offered NO visual cues of an impending continuation of the crash.

2.It perfectly nailed a re-entry into the JSE on 27th February 2009, a few days before the major trough of 3rd March 2009 - absolutely incredible timing!

3.It got out just in the nick of time in the 1998 crash, selling just after the absolute peak.

4. Whilst there were not "crashes" in the first two periods the LBYC model "sat out" the JSE, the returns the strategy earned in fixed interest (measured as the prime rate less 4%) out-performed the JSE returns in these periods. By definition these are "high interest rate" periods.

5. Whilst the LBYC timing strategy could be accused of "chickening out" too early in the 2002-2008 bull market, it actually held on a little longer than SUPERModel. To be honest, the period between when our yield curve inverted and the ultimate frothy peak of May 2008 was not driven by fundamentals but by greed typified of the "blow-off" period in bull markets. Nonetheless, like SUPERModel, the strategy was earning fat interest safely tucked away in its money market account (as rates were increasing all the way into the peak) and avoided the 3rd most devastating crash since 1929. It then went on to make up for its chicken-heart by getting into the market when doom and gloom was at its peak!

6. Due to the rapid nature of corrections and the slow build-up to peaks, the investor deploying the LBYC timing model would not have to wait long between trades before opportunity presented itself again. For this reason the investor should take care not to stuff his/her cash away in long-term call deposits to try and squeeze an extra 0.5% of interest during the 25% of the time this strategy sits in cash.

7.The LBYC Timing Model would have provided for very efficient tax and dividend returns, due to the long holding periods.

8. The LBYC Timing Model delivered 4.5 times the performance of the JSE buy-and-hold strategy. But even more incredible is the fact that the out-performance gains were attributed entirely to the 25% of the time represented by the 5 periods it "sat out" the market. It got out near the peak, went into cash and earned heaps more than the JSE until it saw another BUY signal and promptly went and loaded up on heaps more JSE stock at bargain basement prices. Just 5 instances of this sell-high, earn-high-interest, then load-up-on-the-low levering effect is what enabled this incredible feat.

Naturally, the first thought that came to our minds was if we could incorporate LBYC into the SUPERModel as another component to enhance its returns. Alas, try as we might the incorporation proved entirely fruitless. The SUPERModel is a sub-set of the LBYC vested periods and already achieves such high returns that the incorporation of LBYC had negligible effect. When LBYC was out the market you could be guaranteed that the SUPERModel was out as well.

One of the reasons is that the LBYC is generally the last of our investing timing models to leave the building and switch off the lights. It scored second in gutsiness in the run-up to the 2008 frothy peak only to the Econometric model, but as we stated in previous research this seemed to be the only instance the Econometric model hung on for dear life.

Actually we are quite pleased, since we have no desire to boost the already incredulous returns of the SUPERModel (its bordering on unbelievable already) and frankly we welcome another investment timing model that stands on its own in its own little niche as this assists us with timing strategy diversification. We now have two successful long-term timing models, each based on an entirely different set of indicators which provides for excellent diversification. If one timing model should for whatever reason be unsuccessful chances are the other will be fine.

What we recommend is that in a similar fashion we diversified between various signal scores of the SUPERModel in "The PowerStocks Way" (where we phase in and out percentages of our capital) that you also concurrently allocate a portion of your funds to the LBYC timing model as an insurance you won't be left out of the last bit of a bull market when our other models have headed for the hills. At least you have the comfort that the odds are on your side you will be kept out of any non-"Black Swan" crashes related to the economy or recessions! By way of example, we recommend that the 45% of your capital we suggested in "The PowerStocks Way" that you allocate to SUPERModel be reduced to 30% and you allocate the remaining 15% to the LBYC strategy.

It is not entirely infeasible that the LBYC model could deliver a SELL signal before the SUPERModel or any of its components. In the past it has happened on 1 occasion. And that brings us to another useful purpose of this instrument, even if we have not applied any capital to it - when it says SELL, best we look really hard at what's going on around us - it's the sell signal of last resort!

Undoubtedly, this timing model will become an important part of the PowerStocks Investment timing toolbox for our ENHANCED/PRO subscribers and as from 15th January it will be put into production alongside SUPERModel as part of our long-term timing arsenal.

We will publish, on a once-per week basis in The Weekly JSE Pulse a LBYC signal chart that will represent the yield differential (spread) between our two baskets of long and short term debt of similar risk, centred around our specially calibrated "neutral line" representing zero less the market reward for interest rate risk at the time, to act as a BUY/SELL trigger.

The signal as it would appear over the last 26 years, and similar to how it will be displayed to our subscribers in 2010, is shown below:

It is interesting to note the hilly-shaped outlines of the differentials and how they form definite peaks before beginning steady, irrecoverable declines that eventually signal the end of the bull market. This is very useful information for the market-timer as it gives clues to how much "oomph" is left in the bull market and as our subscribers will note later when we launch the service, we have developed a set of very reliable tools that can use the curve shape to forecast expected end-dates of the bull market (not the returns though, now that would really be something wouldn't it?). Similarly the behaviour of the curves when in negative territory offer "some" clues as to how much longer a recession or bear market has to go (but less reliably than the positive curves.)

We have programmed in a black line dubbed "Normal Curve" and anything above zero up to 2% spread is regarded as "Normal Yield Curve" territory. Anything above 2% we regard as "Steep Curve" territory. Close to zero represents "Flat curve" territory and the spread dropping below this is our SELL signal. Anything below zero represents an "Inverted yield curve". Remember from a previous paragraph:

1. If the yield curve is inverted, the investor should not buy long-term assets and stay liquid.
2. If the yield curve is flat he/she is indifferent, but would probably be best off staying liquid
3. If the yield curve is normal he/she is indifferent but would probably be best off staying invested.
4. If the yield curve is steep he should definitely be invested.

Now a final observation before we close - look at the yield curve spread during the recent recovery of the JSE. Note how we are in "Steep Curve" territory. But also look at how the slope is starting to flatten out - hinting at the fact that the fat party we just had may be nearing an end. Not a sell signal, merely a sign that we cannot expect the same returns we have since March 2009. When the slope of the spread picks up its skirts again its a sign to us another run is coming for the JSE. The direction and angle of the spread curve in our chart is going to be important cues to PowerStocks subscribers.

The Long Bond Yield Curve Spread is a powerful, reliable and accurate forecaster of pending downturns in the SA economic cycle and indeed provides for excellent short-term warnings of major JSE corrections.

You do not want to be dabbling in the JSE nor have any capital invested in the JSE when the LBYC signal is below zero (an inverted yield curve)

Using the LBYC as a long-term investing market timing signal together with an ETF such as SATRIX40 (or any other ETF that proxies the JSE) has provided, historically, 4.5 times the returns of a JSE buy and hold strategy by being vested in the JSE 75% of the time.

The returns gained from an investing strategy centred around the LBYC timing model is tax efficient and has low likelihood of portfolio disturbances likely to attract personal gains taxes.

The PowerStocks LBYC Spread Timing Signal Chart provides visual confirmation of when to be in the JSE and when to not, and also provides clues as to the size of expected future returns from the JSE based on the direction and slope of the signal line, and the position of the line relative to the "Normal Curve" demarcation. It can also be used to estimate "time to live" of a bull market.

Whilst the LBYC has an excellent 100% prediction rate of corrections when it fires a SELL signal, the lack of a sell signal is no guarantee no major correction could occur as shown by the "Black Swan" event of the crash of 1987.

Since LBYC is a "reluctant seller of equity" it hangs onto the JSE for dear life and as a result the investor is likely to encounter draw-downs of up to 15% from time to time during the normal gyrations of the JSE. However the investor is "relatively safe" in the knowledge that the chances of these corrections being major crashes is statistically very low.

The LBYC timing model provides PowerStocks subscribers with an excellent vehicle for long-term investing timing model strategy diversification when used together with the SUPERModel Strategy since both timing models use entirely different indicators and metrics to come to their recommendations of being in or out the market. It is suggested that of the 45% of your capital we recommend in "The PowerStocks Way" that investors allocate to long term investments, that 30% be allocated to the SUPERModel and 15% to the LBYC strategy.

On the very improbable occasion the LBYC should cough-up a SELL signal before SUPERModel, we had best sit up and take notice of what is going on around us.

OTHER INVESTMENT MODELS --> | SuperMODEL | The Big Mac | Cash/Equity Yield Model |
Make a Free Website with Yola.