Seeking Alpha

Kurtis Hemmerling's  Instablog

Kurtis Hemmerling
Send Message
I am a quantitative model designer. I build customized portfolio solutions for brokerages, family offices and individual clients around the world. Some of the models required the capacity to trade hundreds of millions of dollars. I would best describe myself as a value investor looking for... More
My company:
My book:
Myth-Busting the Stock Market: Developing Uncommon Sense By Challenging Financial Truisms
View Kurtis Hemmerling's Instablogs on:
  • Welcome To Transcendent Stock & ETF Portfolios

    In May 2015 I launched a new subscription service on Seeking Alpha called Transcendent Stock & ETF Portfolios. What is this service and what can a subscriber expect to receive for $235 annually?

    There are 5 portfolios which are updated every 4 weeks. They are as follows:

    1. Smallcap Pairs Trading Pro
    2. High Yield Income Plus
    3. Quality Dividend Growth
    4. Asset Allocation ETF System
    5. S&P 500 Turnaround Stocks

    As well, I provide subscribers with my current market-timing sentiment. Let's look at these portfolios one at a time.

    Smallcap Pairs Trading Pro

    This is a market neutral strategy. This means that you can potentially generate positive returns whether the market is going up or down. You are to risk equal amounts of capital on both long and short sides of the market. You buy 2 stocks and short sell 2 stocks in each sector.

    How does this system select stocks to buy and short? It ranks a stock based on various criteria relating to sentiment. For example, it will look for any change in analyst sentiment. Have future earnings forecasts been revised upwards recently? Has the average analyst recommendation improved? What about institutions - how much stock do they own and how much of the available shares are they currently purchasing? What does short interest tell us about investor sentiment?

    This chart outlines the in-sample performance from 1999 until the portfolio went live on May 4th 2015.

    (click to enlarge)

    This next chart shows the out-of-sample performance from May to October.

    (click to enlarge)

    High Yield Income Plus

    Investing in high yield dividend stocks is a common practice - one that can assume a lot of risk and be very dangerous unless you know what you are doing. Stocks offering large dividend yields are usually priced low by investors for a reason. The difficulty lies in determining if that reason is valid or not. Are the yields large simply because the broad market is forecast to have superior growth than these lower growth companies? Is there panic selling taking place? Is the company in distress?

    The High Yield Income Plus portfolio analyzes various valuation factors and combines this with a little bit of positive momentum to isolate strong performing dividend stocks. 50 stocks are presented every 4 weeks in an ordered list to show you which is the highest ranked ticker. A popular diversification method is to choose the top 2 or 3 stocks from each sector in the list.

    From 1999 until May 4th 2015 when the portfolio went live, the in-sample performance looks like this:

    (click to enlarge)

    The out of sample performance from May to October 2015 looks like this:

    (click to enlarge)

    The 2 charts above represent the performance of the entire list of 50 high yield stocks. A superior risk/reward profile can be achieved through the diversification method mentioned above.

    Quality Dividend Growth

    An increasingly popular practice is emerging called dividend growth investing. Investors buy stocks which have a minimum number of years of consecutive dividend growth. The idea is that the growing dividend pot can be forecast to meet future income needs. Yet, what makes one stock with 10 consecutive years of dividend growth a safer bet than the next one? The answer is quality.

    Based on a research paper, the heart of this system centers on separating quality dividend growth stocks from junk stocks based on a quality ranking system. 4 key areas are analyzed which include profit, growth, safety and payout. You can read more about these in the white paper found here.

    This is the performance of the 20 stock dividend growth portfolio from 2000 until May 2015. I use the year 2000 since more historical data is necessary to calculate a 10 year increase of dividends.

    (click to enlarge)

    This is the out-of-sample performance from May to October 2015.

    (click to enlarge)

    Asset Allocation ETF System

    An important investing concept is diversification. An investor who divvies up his capital between various asset classes often has less volatility. One class may go up while another drops thus canceling out each other. How can you allocate capital between classes to lower risk?

    One method uses long-term moving averages. Risk of a downside drop increases when prices fall below the 200 day, or 10 month, moving average. This asset allocation system will only purchase asset funds which stay above the 200 day moving average. As well, if the S&P 500 ETF is held (NYSEARCA:SPY), a diversification rule will not hold another fund which is highly correlated to it. This protects the investor in the event of a highly correlated crash between asset classes. Finally, funds are ranked based on the strength of the uptrend price movement. The 4 strongest asset classes are held.

    From 1999 until May 2015:

    (click to enlarge)

    From May to October 2015:

    (click to enlarge)

    After a sharp downturn in the market, the Asset Allocation system went defensively to 75% cash and 25% into short-term bonds. We will have to wait and see how this defensive play turns out in the coming months.

    S&P 500 Turnaround Stocks

    Investors love to buy stocks rebounding stocks. This screen takes stocks which have under-performed but are poised to take prices to the next level. This portfolio analyzes big cap valuation looking for the best deals as well as various sentiment indicators to signal when the tides are about to turn.

    From 2000 to May 2015:

    (click to enlarge)

    From May to October 2015:

    (click to enlarge)

    Market-Timing Indicator

    Every week I track the market using this simple but effective market-timing technique. Three variables are tracked and when any two of the three turn negative, I encourage caution. The indicators are:

    • S&P 500 earnings estimates
    • S&P 500 price action
    • The US unemployment rate

    When forecast earnings turn downward, this often starts a trend of panic, a drop in consumer spending followed by an even deeper downward earnings revision. When the trend drops - this signal turns negative. The same is true of the S&P 500 price action. Finally, when unemployment rates spike, a downturn is often not far behind.

    This chart shows how the market-timing signal has worked since 1999. When the signal is negative, no positions are held. When the signal is positive, all S&P 500 stocks are held equal weight.

    (click to enlarge)

    Of course, I do not encourage this type of all or nothing market-timing. My recommendation is that when the market-timing indicator is negative, follow any regular sell recommendations that your model portfolio suggests. But do not initiate any new positions that the model recommends until the market-timing indicator is once again positive.


    These are the 5 models currently offered through the subscription service here on Seeking Alpha. Please feel free to ask me any questions you may have or make suggestions for future models that I can offer.


    Kurtis Hemmerling

    Oct 09 6:53 AM | Link | Comment!
  • S&P 500 Market Timing - September 29th 2015

    I don't believe that anyone can perfectly time the market. With that disclaimer, I also believe that the market gives various indications before a bear market hits. Sometimes you can discern the signals in advance - other times you cannot. Even if you are correct, knowing how to act on the signal is another story. What are the signals telling us right now?

    My Simple Market-Timing System

    The market-timing signals that I use are fairly simple. I will have a negative sentiment if any 2 of these signals are true. If less than 2 signals are true, my sentiment remains positive.

    • The S&P 500 price is trending downwards
    • The aggregate S&P 500 earnings forecast is trending downwards
    • The unemployment rate in the United States is trending upwards

    On September 21st 2015, I alerted investors who subscriber to the Transcendent Stock & ETF Portfolios service on Seeking Alpha that the market-timing signal had turned negative. What conditions led to this call?

    Condition 1: Negative Price Action

    First, let us tackle the obvious one. The S&P 500 index price has been trending downwards for some time now. Look at the chart below.

    (click to enlarge)S&P 500 Daily Price Chart

    The first condition was met when the daily price for the S&P 500 fell below its 100-day moving average - which is the blue line. This happened in August.

    Supplementary: You may also notice the trend-lines which are the parallel upwards sloping green lines. Although this isn't included in the market-timing signal, I find it very useful to draw long-term trends above and below the price action. A downtrend usually has a sharp downward break of the bottom trend-line which is often followed by a small upwards rally. Sometimes the rally succeeds but other times the price hits the underside of the trendline, waffles, and then drops down hard.

    My stance would turn particularly negative if the S&P 500 ETF (NYSEARCA:SPY) fell below 182, a level which is providing some price support. But this is my personal bias and not really part of the market-timing signal.

    Condition 2: Negative Earnings Estimate

    The second criteria considered is the S&P 500 earnings estimate. Below is the aggregate analyst earnings estimate(current year) for the S&P 500 since 1999.

    (click to enlarge)S&P 500 EPS Analyst Estimate for Current Year

    I will zoom in on the recent estimation of the S&P 500 EPS.

    (click to enlarge)Market-Timing Signals

    I filled in the green circle at the beginning of July where the short-term moving average (5 week) EPS estimate crossed above the medium-term moving average (21 week). This suggests an optimistic trend where earnings are being revised upwards. The signal crossed back down for a negative sentiment on September 19th.

    This second negative condition flipped my sentiment to bearish.

    Condition 3: Unemployment

    When unemployment rates spike, the market often drops. This US unemployment chart clearly shows that since 2010 the rate has been steadily dropping. This is a bullish indication. However, the other 2 bearish signals create an overall negative sentiment.

    USA Unemployment Figures

    How I Use Market-Timing

    Few people have the desire or the will to go 'all in' or 'all out' of the market. If you have ever missed a rally because your indicator was wrong - you probably won't be making that mistake twice. As well, transaction fees and slippage can erode your capital if you over-trade market-timing signals.

    I prefer to employ 'soft market-timing'. Unless you are a buy and hold investor, you should have buy and sell criteria. Every week, review any potential buy and sell recommendations. During periods of bearish sentiment, I execute the sell recommendations but not the buy recommendations. This slowly converts your portfolio to cash in the event of an extended downturn, yet you are still partially invested if the market reverses. As well, you do not need to sell all your holdings at once - just the weakest ones on a weekly basis.

    Here is an example of one portfolio compliments of Portfolio123:

    (click to enlarge)Portfolio123 Stock Screening and Backtesting Chart

    The top window represents portfolio performance from May 2008 to 2009. The second window shows maximum portfolio drawdown - which is the maximum loss when counting from the previous high. The third window shows the amount of capital invested, which goes down as the portfolio raises cash when the market-timing indicator is bearish.

    Market-timing is not for everyone. Even those willing to experiment with this practice may prefer a compromise of progressive selling, such as outlined above, rather than an all-or-nothing signal that is rarely acted on.

    Sep 29 2:37 AM | Link | Comment!
  • FAQ For High Yield Income Plus Portfolio

    This blog is to answer a few questions regarding one of the models offered in the new SA Premium program - High Yield Income Plus.

    The article, Overcoming The Pitfalls Of High Yield Investing, has raised a few questions in the comment section that deserves further treatment.

    How much of the annual returns is due to dividends vs. capital appreciation?

    Of course, this will fluctuate every year depending on such factors as opportunities for firms to re-invest, the interest lending rate and so forth. Using one of Portfolio123's tools, I create a custom data series that gives the average dividend yield for the universe I select stocks from based on my ranking system.

    (click to enlarge)Avg Dividend Yield High Yield Stocks

    Thus, the average annual dividend yield would be between 9 and 10%. The total return CAGR is over 21% for the 50 positions. We can say that a little less than half of the stocks return is given in dividends, assuming that our 50 stocks have a fairly equal distribution in the stock universe.

    Does market risk related to income risk? If I buy and hold can I ignore market conditions if I am only concerned about the income stream?

    Some investors claim that income risk and market risk have little cross-over. Is this the case? How would we test it?

    First we define our investable universe. In this case we say that our 'high yield' investable universe is any stock that meets the following requirements:

    • Average price>$1 and daily turnover>$50,000
    • Minimum 3% dividend yield
    • Dividend yield at least 1% higher than 20 year Treasury bond yield

    I will simulate buying and holding any stock that passes on the date January 2007. I will then track the total dividends paid (not per share but total $ amount) to common shareholders over the trailing 12 months. (1 = $1,000,000). The chart will be market-cap weighted since we simulate owning 100% of all market-capitalization.

    If income risk is unrelated to market risk, the rising and falling of dividends paid ttm will have low correlation to what the market was doing. To put it another way, rising markets will not necessarily see a rise in dividends and a market crash will not necessarily see a drop in dividends.

    (click to enlarge)

    The data is a little lagged since we are using the trailing 12 months. But it is fairly obvious that the market drop in 2008/09 had a major impact on our dividends even though we bought and held since 2007. The market dropped by over 50% and our annual dividend stream dropped by over 50%.

    How much of the income risk is related to companies no longer being listed?

    The following chart shows the survivorship of our 212 stocks which we invested in starting in 2007.

    (click to enlarge)

    Interestingly enough, there is a steady decline. Thus, our stock universe has an annual discount rate of 5.2% (we lose 5.2% of our stock universe every year).

    This gradual decline should likewise create a gradual drag against our dividend chart in the previous question. Thus, the up and down volatility of total dividend income is not due to delisting stocks.

    To negate the effect of delisting stocks, I will track the average dividend cash flow paid to common shareholders (NYSE:TTM) instead of the sum. Thus 100 stocks and $100 million in total dividends will produce the same ratio as 50 stocks and $50 million in dividends in the event that 50 stocks fall off the market.

    (click to enlarge)

    After controlling for stocks which delist, the remaining stocks still saw a 45.5% reduction in income during the 2008/09 crash.

    Yes, but this is clearly due to financial stocks. What is the result if I did not invest in any financials?

    First of all, it would be incorrect to remove financial stocks in hindsight. Unless you accurately forecasted the market crash due to financials and stayed away from that sector, it needs to stay in our population.

    But for arguments sake, let us remove financials to see the average dividends paid of surviving firms since 2007.

    (click to enlarge)

    Our dividend draw-down is up to 48%. Removing financials did not improve our average income reduction of surviving stocks. While only 4.2% of our stocks drop off the list annually as opposed to 5.2% when including financials, our actual income volatility is greater despite the reduction in stocks dropping off.

    Relating to the very first chart in the linked article at the top, is the 90%+ drawdown of the 25 highest yielding stocks in the Russell 3000 is due to mortgage backed instruments?

    Let us re-run the test removing sub-sector Real Estate (GICS 4040). This test will take the 25 highest yielding stocks in the R3K and then only hold those tickers which are not part of the real estate sub-sector.

    (click to enlarge)

    An average of 11 stocks of the 25 highest dividend yielding names relate to real estate. If we remove those names, the remaining tickers held equal-weight still produces a draw-down of 84%. True, this isn't a 90% drawdown but it is still very ugly and much higher than the broad market (which includes real estate stocks).

    May 21 11:31 PM | Link | Comment!
Full index of posts »
Latest Followers


More »

Latest Comments

Posts by Themes
Instablogs are Seeking Alpha's free blogging platform customized for finance, with instant set up and exposure to millions of readers interested in the financial markets. Publish your own instablog in minutes.