1. Pressure on the front end of the yield curve - This was caused by safehaven buying and repatriation by Americans of foreign investments looking for a place to park their cash and earn a little yield. There is also the (not so small) issue of margin calls. As the markets came crashing down and as hedge funds fell into the loop of selling to cover margin calls causing further price drops and further margin calls, the need to raise cash as collateral increased. As far as margin goes, treasuries are considered to be as good as cash. What we have seen is a stampede into treasuries by large investors and managers in order to maintain their margin levels and earn some yield in the process (as opposed to posting cash).
2. Long end over supplied - And supply is expected to increase. As the U.S debt increases and the yearly budget deficits grow, the government is expected to try to fund the costs of current and future bailouts with longer dated maturities. They are trying to avoid what is already a big issue in the corporate world - rolling short term debt. With yields under 5% and being dragged lower by Fed rate cuts, it’s a great time for the government to fund its long term liabilities. That being said, rates can’t go much lower and once the Fed runs out of ammo, we may see the long maturity rates move up.
3. Foreign governments may be forced to sell or may voluntarily diversify - Taking cue from Warren Buffett, government debt is expensive right now (rates are low). Some countries may find it necessary to sell their treasuries to raise cash for local stimulus, while others may be looking to diversify at a time of a strong dollar and attractive exit prices. Others, such as China, may be looking to add less long term treasuries to a portfolio already centered on U.S. debt.
4. Risks exist to both the sovereign rating of the U.S. and the stability of the currency - In my opinion, the biggest factor driving the steepness of the yield curve is the financial health of America’s balance sheet. Although many are searching for ANY yield right now, as small as it may be, they are looking at the front end of the curve as further down the line carries increasing risks.

It will be interesting to look back in a year to determine which, if either, analysis (i.e. Adam’s or Bespoke’s) was accurate.

The difference in the yields between the 10-Year and 3-Month US Treasuries is currently 375 basis points and that is near a 10 year high.  Quoting from Bespoke’s article, while an inverted yield curve has been a reliable precursor of impending economic weakness, steeply sloped yield curves have historically been a sign that the market is anticipating strength in the economy (as was the case in 2002 and 2003).

I am neither an economist nor a financial analyst so I do not consider myself qualified to present economic forecasts based on the slope of the yield curve but I do find its steep slope to be interesting.

FJP

In my own portfolios, my Canadian holding is the iShares Canadian Short Bond Index ETF (XSB) and my US holding is the iShares 7-10 Year Lehman Treasury Bond ETF (IEF).

You can view updated performance tables and equity curves for the Canadian and US ETF2X models here and here.

There was a good article in the Globe and Mail yesterday by Tom Bradley. Essentially, the message was that when choosing mutual funds to invest in you should look for fund managers that have a significant portion of their personal investments in the funds they manage. I don’t own any mutual funds but I like the message as it can be expanded to other facets of investing. For example, if you want to follow the stock recommendations of a fellow investor and consider Tom’s message to be important then you should consider joining Covestor. Most of the top rated investors have blogs and you can follow their trades to determine if they follow their own recommendations. Admittedly I have a bias here as I am a member of Covestor.

FJP

Related Stocks:  IEF

FJP

Dividends have been included in the above results.

PSQ, RWM, SH and DOG are single exposure contra ETF’s designed to move one times the relevant index daily in the opposite direction. QID, UWM, SSO, DDM and HXD are double exposure contra ETF’s designed to move two times the relevant index daily in the opposite direction.

I can’t say I suggested that you buy contra ETF’s in June as I normally don’t try to short the market.  As time goes by and my models build a more substantial history I may revisit this policy but for now I just suggest that you get out of the market and go to a money market ETF or short term bond ETF when the ETF2X timer signals that it is time to get out of the market.

The Nasdaq responds best to the US version of the ETF2X timer so it is no surprise to me that QID has had the highest return.  Note that the return on the double exposure contra ETF’s is less than two times the return on the corresponding single exposure contra ETF’s. In the case of QID and HXD, both of these are up more than two times the underlying index.

On a different matter, some of you had trouble reading the screen shot from Covestor in my post several days ago.  I’ll try again with the following which illustrates that I am currently ranked #39 based on absolute return over the past three months.

Covestor doesn’t provide information on the number of investors that I am ranked against but I think it is fair to say that there are thousands of others that I am competing against to be in the Top 100.

Related Stocks:  PSQ, RWM, SSO

In Table 1 I have listed the models and their corresponding ulcer index and maximum drawdown.  I have not included the compound annual growth rate (CAGR) because, regardless of the return provided by a model, I choose not to use it if the ulcer index and/or maximum drawdown is too high.

Table 1 Comparison of Models from 09/05/07 to 11/07/08

Now I will remove from further consideration the models which had a maximum drawdown in excess of 15% and sort by ulcer index.

Table 2 Comparison of Suitable ETF2X Models

The top three models by return are QQQQ/Short QQQQ, QLD/IEF and HXU/XSB.  For now I am not going to consider the QQQQ/Short QQQQ model for my own investments since it is skewed by the once-in-a-lifetime market decline we experienced this year.  That leaves us with the QLD/IEF and HXU/XSB models as the best for my personal investor profile.

Now the key question we have to ask ourselves is “Will the short term bond ETF’s (IEF and XSB) be the place to park money when the timer tells us to get out of the market in the future?”.  I don’t have the answer but if you do, please share it with the rest of us!

FJP

Related Stocks:  IEF, QLD, QQQQ

$100,000 invested in Berkshire Hathaway on August 03, 2006 would now be worth $123,768 but a $100,00 investment in the QLD/IEF model would be worth $222,496. Even though double exposure ETF’s are naturally more volatile than the market, the ulcer index for the QLD/IEF model was 32% lower than that for Berkshire (1.71 vs 2.52).  And get this, the maximum drawdown for Berkshire is 29.5% versus 13.4% for the QLD/IEF model!

The QLD/IEF model was selected because it is the US model that I prefer and I invest my own money based on it.  August 03, 2006 was selected as the start date since that corresponds to the first time the ETF2X signal went long after the Proshares Ultra QQQ (QLD) was introduced.

What if we wanted to consider a longer time period?  Let’s compare buying and holding Berkshire against following the IWM/Cash model from June, 2000 to the present.

The IWM/Cash model provided a higher return, a lower ulcer index and a lower maximum drawdown than holding Berkshire.

So do you think Obama will tap me for financial advice instead of Warren?!

FJP

Related Stocks:  IEF, QLD

Here is another interview from Bloomberg last week. Note his thoughts on how smart Bernanke and Greenspan are!

Slightly better quality video is available here.

In my own portfolios, my Canadian holding is XSB and my US holding is IEF. The HXU/XSB model and the QLD/IEF model appeal to me since each has the lowest ulcer index and the lowest maximum drawdown in its group and both have very respectable CAGR’s.  You can view updated performance tables and equity curves for the Canadian and US models here and here.

I moved my personal investments towards the ETF2X models earlier this summer and this has resulted in me ranking in the Top 50 investors at Covestor based on absolute return for the past three months.

Please feel free to send me a note via the contact page to let me know if my blog has been helpful for you.

FJP

Related Stocks:  IEF, QLD

Several years ago, I was at the final assembly for our local high school to make a presentation on behalf of our company when I asked the principal if I could make an announcement.  Earlier I had decided that if I publicly stated that I was going to give to students that amount of money that was given to me while in university vis-a-vis scholarships and the fellowship then I would have to stick to that commitment.  I told the students that I would personally fund an award of the school’s choice (the usual awards such as highest grade in math, etc. were already sponsored).  The following September I met the principal at his office and he asked me if I would fund the student leadership program.  The school (Exploits Valley High) had developed a student leadership course and each year several students attended either the provincial or national student leadership conference.  I agreed to the request and the rest is history.

This year the students attended a national leadership conference in September hosted by Three Oaks High School in Charlottetown, Prince Edward Island.

In addition to the Leadership Award, starting this past June I personally sponsor the Male Student of the Year award which is presented at the graduation.  By the time I am in my mid-fifties I will have passed along all the money that was given to me while in university.  So now you know - there is more to me than my passion for successful investing!

FJP

Related Stocks:  PSQ, QID, QLD

You can view the equity curve for my HXU/XSB model here.  Feel free to comment or contact me.

FJP

I determined the ten best days for the Nasdaq from June 01, 2000 to October 31, 2008 (corresponding with my equity curves here). The ETF2X timer had us out of the market for four of these days so let’s consider what happened when we were out of the market.  The ETF2X timer took us out of the market on September 19, 2000 and got us back in on January 22, 2001 and we therefore missed a 10.5% increase in the Nasdaq on December 05, 2000. Well, being out the market wasn’t such a bad thing since the Nasdaq lost 28.7% from September 19, 2000 to January 22, 2001.

On another instance, the ETF2X timer had us out of the market from April 26, 2002 to  October 25, 2002 and we therefore missed a 7.8% jump in the Nasdaq on May 08, 2002.  Following the ETF2X timer would have been a good thing since the Nasdaq lost 20.0% from April 26, 2002 to  October 25, 2002.

The other two instances of high daily returns for the Nasdaq while we were out of the market occured last month. The ETF2X timer took us out of the market on June 11, 2008 and the Nasdaq is down 28.1% since then.

In not one single case were we worse off by following the ETF2X timer and being out of the market on one of the best 10 days for the Nasdaq.  The next time you see an article which touts the benefits of a buy-and-hold strategy because of the ill effects of being out of the market on the top ten days, you can smile since you know better.

Top 10 Days for the Nasdaq

When asked what he is doing with his own investments, Dr. Roubini said that he is a long term investor but if he was an active investor he would play the market to the downside.

I have posted updated performance tables and equity curves for the Canadian and US models here and here.  The ETF2X timers didn’t generate new signals this past week but the US timer is getting close to going long.  If the US markets move higher throughout next week, I suspect the timer will go long but it won’t happen before Thursday.  We have experienced a rally lately but many pundits have classified it as a relief rally.  The ETF2X timers are designed to not act too fast thereby preventing its users from getting whipsawed with ever-changing signals.

I don’t hold any short (contra) ETF’s such as QID and RWM in my US account but if I did I would sell them and hold cash. You should have a plan for the actions you intend to take if the ETF2X timer goes long for the US market.  Personally I am considering buying QLD and IWM.  Although I have mixed feelings about stops, I may use 12% trailing stops for half of my purchases.

Please feel free to send me a note via the contact page.

Related Stocks:  QLD, RWM

Brooke’s model is based on buying on October 28 and selling on May 05 the following year. Both the buy-and-hold model and Brooke’s model are based on the Russell 2000 index.

I ran the numbers twice to determine if this year’s market meltdown affected the relative performance of the three strategies.  In both cases the compound annual growth rate (CAGR) is improved by both selecting the October 28 to May 05 holding period and by using the ETF2X timer.  For my own sake, it was reassuring to see that the ETF2X timer gave the best results.  If you want to see an equity curve of the IWM model I have developed, please click here and scroll down the page until you see it.

Where Brooke and I differ is our views on forecasting the market.  I will not forecast a move in the market in either direction until my timer changes its signal.  Brooke, however, in last month’s market letter advised investors to get into the stock market yesterday on the premise that it was the start of a traditional six month favourable period.  Brooke notes that the S&P500 has provided a positive return from October 28 to May 5 79% of the time from 1950 to 2007.  If my timer generates a long signal soon, I’ll be very comfortable knowing that seasonal history is on my side at least until May 5!

This month’s market letter contains some interesting factoids about the Dow Jones Industrial Average.  For example, the DJIA spent 40 years range bound around 100, spent 20 years range bound around 1,000, and now appears to be range bound around 10,000.  Brooke believes that there is a strong likelihood that the US market will trade sideways for a number of years.

You can sign up for Brooke’s free monthly market letter here.  I am a recent subscriber to the Thackray Market Letter and I like what I have seen so far.  There is no sales pressure to buy other products and Brooke doesn’t appear to be a talking head for a financial services firm (besides his own of course).

Canadian Models

US Models

If you are wondering how I have done personally since following the ETF2X timers starting this past summer, the following Covestor chart shows how my Canadian holdings (cash excluded) have performed relative to the S&P/TSX composite index.

After each new signal, a 15% trailing stop is set up. If the trailing stop is triggered then the model goes to cash until a new signal is generated.

Below are equity curves showing the QQQQ model with no trailing stop and the same model with a 15% trailing stop.

The net result of applying the trailing stop is an improvement in the compound annual growth rate from 11.1% to 12.7%. Not a drastic improvement but there is no reason to not use a trailing stop as part of your money management rules.

I could test the model for various trailing stops (e.g. 10%, 12%, 14%) to find the optimum stop but then I would be guilty of optimizing the model based on backtesting which is not my intent.

I investigated whether a trailing stop would be beneficial to my long only IWM model but the model has such a smooth equity curve that there is no benefit to using a trailing stop.

Related Stocks:  QQQQ

I thought it would be interesting to try another Powershares QQQQ model for which instead of going to cash, it short sells QQQQ’s.  The resulting equity curve is below.

This model has a higher compound annual growth rate than the QQQQ/Cash model which has obviously been greatly enhanced by this year’s market meltdown.  However, the maximum drawdown is 44% which is still much too high for me.  From a cursory review of the numbers, I think a 12% trailing stop would have a positive effect on the drawdown.

Related Stocks:  QQQQ

Powershares QQQQ

Proshares QLD

I am comparing writing a covered call on Powershares QQQQ and Proshares QLD.  Out of interest, I wanted to see what the annualized downside protection is for both ETF’s. Clearly it is much higher for QLD but then it should be since QLD is a double exposure (2X) ETF.  There is one problem with selling QLD calls and that is the tiny open interest (104 compared to 44828 for QQQQ).

I am always thinking about ways to enhance the ETF2X models and writing covered calls is one of the avenues I may pursue.

Related Stocks:  QLD, QQQQ

For the fun of it, I thought I would compare my worst performing US model against buying and holding shares in Warren Buffett’s company, Berkshire Hathaway.  Had you bought shares in Berkshire at the open on August 03, 2006 and held them until today you would have a 30.9% gain.  If, on the other hand, you followed the QLD/TIP model you would have a gain of 88.5%.  Of course commissions are a factor but there have only been nine ETF2X signals since August, 2006.

Comparing only the rate of return doesn’t provide a thorough analysis so let’s consider maximum drawdown and ulcer index.  The maximum drawdown and ulcer index for the QLD/TIP model are 17.8% and 1.47 respectively.  For a buy and hold Berkshire Hathaway strategy, the maximum drawdown is 25.1% and the ulcer index is 2.34.  The QLD/TIP model outperforms on all metrics considered!

On the Canadian front, the HXU/XSB model is still the top performer.

As I moved my Canadian investments toward the HXU/XSB model this summer, I have avoided a large degree of the decline in the Canadian stock market.

Related Stocks:  QLD, TIP

Had you bought and held the Nasdaq-100 on June 08, 2000 and held until October 15, 2008, you would have lost 66%. Had you bought and held the Nasdaq composite index on June 08, 2000 and held until October 15, 2008, you would have lost 58%. If however, you had bought and sold units of QQQQ starting on June 08, 2000 as per the ETF2X timer you would have gained 4%.

The improvement in the compound annual growth rate achieved by using the ETF2X timer is just over 10%. Note that the ETF2X timer is not one of those hyperactive signals that requires you to get in and out of the market several times per month. For many investors, the ETF2X timer should be relatively easy to implement in a successful investment strategy.

Related Stocks:  QQQQ

The equity curve was constructed by buying units of IWM at the open the next trading day after a long signal was generated and selling the IWM units the next trading day after a market neutral signal is generated.  For the purposes of this model, the proceeds from sales of IWM are held in cash with no interest accrued.

The above equity curve covers the period from June 08, 2000 to October 10, 2008.  The compound annual growth rate of the IWM model with ETF2X timers applied is 11.5% compared to 0.2% had one bought and held the Russell 2000 index.  The maximum drawdown for our model was 13.9% compared to 41.7% had one bought and held the Russell 2000 index.  For me, this evaluation provides a level of comfort to using the ETF2X timers.

I don’t advocate attempting to short the market so I present the QLD/QID equity curve below for entertainment only!

Had you invested $100,000 based on the QLD/QID model on 08/02/06, that investment would be worth $431,655 today.  Of course that couldn’t have happened since I started publishing the ETF2X models only recently.

Related Stocks:  QLD

Let’s be honest.  The models have done well in part due to the once-in-a-lifetime meltdown of the markets.  There are many wise investors who refuse to believe that it is possible to develop a successful market timing strategy over a long term due to the degree of randomness in the markets.  After all, a number of peer-reviewed articles have been presented in prestigious publications such as the Journal of Investing which detail how unlikely it is for market timing to be successful over the long term. Trouble is, none of the articles, to the best of my knowledge, have focused on the type of models discussed here at ETF2X.  That is, none of the studies have examined models based on timers that are correct on the long side more than 60% of the time and employ double exposure ETF’s.  There is a good reason for that, however.  Double exposure ETF’s are so new that long term studies are impossible.  In the meantime, subscribers to ETF2X can decide for themselves if they want to put a portion of their investments behind one or more of the models presented on this site.

Claymore 1-5 Yr Laddered Government Bond ETF seeks to provide a return based on the price and performance, before fees and expenses, of the DEX 1-5 yr Government Bond Index (“the Index”). The investment strategy of the fund is to invest in and hold the constituent securities of the index.

The ETF provides suitable investors of different sizes with the opportunity to gain exposure to a well diversified government bond portfolio, designed with staggered (“laddered”) maturity levels from one year to five year.

I like the fact that the Claymore ETF has a very low 0.16% MER.  By comparison, iShares Canadian Short Bond Index Fund (XSB) has a MER of 0.25%.

From the iShares.ca site:

The iShares™ CDN Short Bond Index Fund seeks to provide income by replicating, to the extent possible, the performance of the DEX Short Term Bond Index™, net of expenses. The DEX Short Term Bond Index is a market capitalization weighted index consisting of a broadly diversified range of investment grade federal, provincial, municipal and corporate bonds with a term to maturity between one and five years.

So both bond ETF’s invest in 1 to 5 year Canadian bonds but the difference is that the Claymore ETF buys only government bonds. The iShares bond ETF includes corporate bonds in its holdings.

In terms of performance, from February 02, 2008 to October 03, 2008 the Claymore ETF (CLF) is up by 3.10% whereas the iShares ETF (XSB) is up by 2.75%.  There is one catch that may be of interest - the average volume for XSB is 95,000 but it is only 19,000 for CLF.

My Canadian model performance table will no longer include a HXU / Cash model since this model is based on putting cash into a non-interest bearing account which doesn’t make sense now that we have a money market ETF model (i.e. HXU / CMR).  Starting with the next ETF model update, the Canadian table will include a HXU / CLF model.

Related Stocks:  MER

I moved my Canadian portfolio towards the HXU/XSB model earlier this summer and below is a screenshot from Covestor showing how my portfolio has done relative to the S&P/TSX composite index.  Please note that Covestor does not track cash held in a portfolio - only stocks, ETF’s and mutual funds.

The ETF2X Canadian timer signalled that it was time to get of the market on June 12.  Since then the S&P/TSX Composite Index has lost 25.9% of its value.

If you held your stocks and ETF’s through the market decline, just think about how much you would have saved by acting on the ETF2X timers.  I am in no way stating that I knew the markets would decline as much as they have.  However, from the back-testing that I have done with the ETF2X timers I believe that taking direction from them beats a buy-and-hold strategy hands down.

Using a proven timer to get out of the market once a downtrend has started will help you work towards Warren Buffet’s two rules of investing:

1. Never lose money.

2. Never forget Rule #1.

For the ETF2X models based on Horizons Beta Pro ETF’s, the HXU/XSB model, with a 26.0% year-to-date gain, is slightly ahead of the HXU/CMR model. Compare these models’ returns against an 11.3% year-to-date loss for the S&P/TSX 60.

Related Stocks:  QLD

I didn’t have a bad day though.  Most of my holdings are in XSB which gained 0.50% during the market meltdown.

My Proshares QLD/QID model is too volatile for me but for any of my subscribers who do follow it today was a great day since QID jumped 20%.  The QLD/QID model now has a compound annual growth rate of 72.4% since August 02, 2006.

Related Stocks:  QLD

CAGR stands for compound annual growth rate and you can get a definition of it at Investopedia.

Ulcer Index is a risk measurement tool developed by Peter Martin in 1987. You can read about the Ulcer Index at Wikipedia.

Related Stocks:  QLD

Related Stocks:  QLD

The cost for a "Platinum Subscription" to Stock Traders Daily Service is $199/mo.  This subscription offers more than market timing but I'll  leave it up to my subscribers to decide where they get the best value.

Related Stocks:  MO, QLD, TIP

The fund invests primarily in high-quality, short-term (generally 90 days or less) debt securities, including treasury bills and promissory notes issued or guaranteed by Canadian governments or their agencies, bankers acceptances and commercial paper (excluding asset-backed commercial paper) issued by Canadian Chartered banks, loan companies, trust companies and corporations.

The Claymore Premium Money Market ETF will have one of the lowest expense ratios of any money market mutual fund in Canada. By providing a lower Management Fee, the fund can reduce risk in the underlying holdings, while generating attractive short-term income rate.

My rationale behind using a money market ETF is that there is no point in parking cash in a non-interest bearing account as is currently the case with the HXU/Cash model.  Oddly enough, the money market ETF is available in Canada but there is no liquid equivalent in the US yet.  Therefore my US models will continue to include a QLD/Cash entry.

The average volume of CMR traded daily is around 12,300 ($615,000).  The Claymore Premium Money Market ETF is less than one year old (incepted on 02/19/08) so there are no performance results on Claymore's site yet. The management fee for this ETF is 0.25% and the current distribution rate is 2.67%.

Related Stocks:  QLD

As for the US markets, the Nasdaq fell 2.64% today.  My QLD/IEF model is one that I am comfortable with even though it has a lower CAGR than the QLD/QID model (43% vs 55%).  The unsettling characteristic of the QLD/QID model is its maximum drawdown of 36% as compared to a maximum drawdown of 16% for the QLD/IEF model.

Related Stocks:  QLD

I want to test this idea to determine if it will improve my timer but I can't find a site that provides minute-by-minute charts for time periods beyond the previous five trading days.

The only good news for me from this past week is that by following my HXU/XSB model, my Canadian investments have beaten the market over the past three months.  By holding XSB units, I have avoided the market decline.

Related Stocks:  QLD

Related Stocks:  QLD

I plan to refer to two books that will assist in my trading plan development: Come Into My Trading Room by Dr. Alexander Elder and Trading In The Zone: Master The Market With Confidence, Discipline and a Winning Attitude by Mark Douglas.