It Has Gotten Worse

In the previous posting we called attention to the call for 40% fixed, 60% equity in July.  What does a 60/40 allocation look like as of Jan 1?

It now projects 20% more downside risk than upside potential if only passive indexes are used.  Even selecting a combination of active ETF’s has 10% more downside risk than upside potential.

What if we go to 50% fixed?

Now the UP Ratio is 14 times more upside potential than downside risk with a 15% allocation to Vanguard Growth Technology, 8.9% allocation to the Hong Kong ETF, 7.8% to the MidCap Growth iShare ETF and 4.7% allocation to the iShares Asia X Japan ETF.  This is not a well diversified portfolio, but it is close to mine.  I overweight China Technology and the Spyder, which over weights Technology.  I use the model as a guide and my brain as the decision maker.  I am currently at 50% Equity and I plan to move to 40% fixed in the very near future.

This is how I recommend professionals use this model.  Amateurs should not use it.




Posted in Uncategorized | Leave a comment

The Roller Coaster Market Ain’t Over

                                      Where did PMPT say we were in July 2018?

The graph below shows the effect of increasing equity above 60% (left graph).  The upside potential ratio for the active portfolio with a 6.2% allocation to VGT, the Vanguard Growth Technology ETF becomes negative, i.e., there would be more downside risk than upside potential. By increasing equity 10% you would incur 30% more downside risk than upside potential.  The passive portfolio (no active managers) would fair even worse, it would drop from an even trade off to 50% more downside risk than upside potential.

The question is, does the theory developed at PRI, before the tech bull market and the Trump election, still provide useful information? Has risk increased or decreased in the past decade?  PMPT said you are not getting paid to take more risk than the 60/40 market mix.  If you have exceeded that magic mix of bygone years, put some stop loss orders in to reinstate a 60% equity position.

It turns out, that was good advice then.

What Now?

As stated in the  Article below I started investing in equities again in November and increased equity to 90% in late December.  I now have a 3.5% profit in my portfolio after   I got stopped out of NVDIA for a small profit and sold CRM for a 20% profit.

What Comes Next?

Next week’s meeting with China should provide a continued rally.  I will raise my stop loss orders accordingly.  Why? the Federal debt, the corporate debt, the junk bond debt, the BBB debt that is really junk, the money market debt that is full of junk.  Proceeds have gone and will continue to go into short term treasuries.  It’s a beautiful day for a family drive in our Tesla to enjoy the good life.

Posted in Uncategorized | Leave a comment



I have now sold my stock in Tesla at a ridiculous profit.  Still have the car and love it!


It has been a long time since my last posting on Elon Musk.  Since then I have bought a model S Tesla and Tesla stock.

Posted in Uncategorized | Leave a comment

On Embracing Complexity

Recently deceased scientist, Stephen Hawking, said to a graduating class, “The next century [21st] will be the century of complexity” and encouraged them to embrace it.  Yes, we should not fear it, but embrace it by gaining knowledge of a new science, Artificial Intelligence (AI).  It began at Dartmouth University in 1956 when John McCarthy used the term (AI) as the title of a conference. To help us operate successfully in an increasingly complex world, we need to learn the language of complexity.

Why have so many experts who forecasted a market top in the past few years been wrong? Perhaps it is because they have been using 20th century methodologies in a 21st century world of increasing complexity.  This is no longer like us playing a game of chess against a few experts and a lot of amateurs, as existed in the 50’s and 60’s.  We are playing against computer programs that have learned to reason from massive amounts of data derived from a very complex system, and are making better decisions than us about everything from driving cars to investing in the stock market.  That means we need to know something about complex systems and the computer programs that understand them better than we do.

The evidence:

The 1st shot across the bow was when IBM’s Deep Blue beat the world champion chess player, Gary Kasperov, because it knew all the games he and every other expert had ever played and could access them in a flash.  That was the earliest form of AI  which involved 3 steps:

  1. Training the model on a massive amount of data shoveled into computers run in parallel on processing platforms.
  2. Testing the model to see if it learned how to win in the past.
  3. Validating the model to see if it can uncover relationships from the data to predict future outcomes, then, learn from and adjust to future events better than human experts.

Another breakthrough came in March of 2016.  The Google program, named AlphaGo, was taught to play the game of Go that has 10 to the 127th power possible moves.  AlphaGo scored a decisive win over Lee Sedol, a top-ranked international Go player, winning 4 out of a 5-game series in South Korea.  Then, in 2017 it won the first two games of Go in a three game championship game with the world’s champion, Ke Jie. The Google AI program appeared to think and reason better than human experts. Below is an example of a supervised learning approach which requires a massive amount of historic data to teach the model how to achieve the goal, for example, beat the S&P 500.  This involves some of the language you need to be familiar with.

Neural Network (Supervised Learning)

The Xi are data sets affecting the thing you are trying to forecast, like, turning points in the stock market.  The Wi are the initial weights assigned by experts based on historical results. For example, X1 could be the 3 year moving average return on the S&P 500.  X2 could be the 3 year moving average standard deviation.  X3 could be the upside potential ratio calculated from the first two Xi with the aid of the Forsey-Sortino model.

The tree search in AlphaGo then evaluates positions using statistical methods like Logistic Regression and Random Forests to adjust the weights. The model then makes a forecast and learns from the results how the weights should be adjusted further. This is called reinforcement learning.

The Break Through (Unsupervised Learning)

In 2018, Google updated the AlphaGo model to AlphaGo Zero, which uses an algorithm based solely on reinforcement learning, without human data, guidance, or domain knowledge beyond game rules . This is known as Unsupervised Learning. AlphaGo Zero became a neural network trained to predict AlphaGo’s own moves without requiring massive amounts of historic data. It learned as things changed in real time. How did it do? In 4 days the Zero version learned how to beat the original AlphaGo model in 100 out of 100 games.

What might somebody using an AI model see that you and I don’t see?

Similar to the AlphaGo Zero model mentioned above, AlgoDynamix utilizes an unsupervised machine learning algorithm to process limit buy and sell orders from financial exchanges to identify patterns as shown above. It seems there are certain patterns of buyers and sellers that the AI model has identified in clusters which have predictive power, the rest are noise.  To be sure, I know very little about this company or the methodology they are using.  However, the result shown below leads me to want to know more.

For more information visit The few graphs in this paper were selected from an introductory course in AI at Stanford University and do not begin to explain the revolution akin to the industrial revolution that is taking place.  I for one intend to find out more about the research being done in AI so I can comfortably surrender my 20th century driving skills to my new Tesla Model 3 when it rolls off the assembly line.  I encourage you to go to your nearest university to take some introductory classes.  You may want to start by downloading:  Kaplan, Jerry. “Artificial Intelligence: What Everyone Needs to Know”. Oxford University Press. Kindle Edition.


It is estimated that AI will have replaced 13 million U.S. workers by the end of 2018 and 1 billion workers worldwide by 2022. Most vulnerable: Truck & Uber drivers, retail workers and… investment advisors ?

“Last year, On March 28, the CEO of BlackRock sacked seven fund managers and shifted billions of dollars to an AI product called Systematic Active Equities (SAE). SAE now employs 80 portfolio managers and researchers, who include more than 30 PhDs in computer science, physics and engineering. This is one of the greatest opportunities in active management that I have ever seen, said Ron Kahn, head of research at SAE.”

Is BlackRock going to need RIA’s to sell their product? Are RIA’s going to need firms like BlackRock to keep their clients? How do they do that if they don’t understand what BlackRock is doing vis-a- vis AlgoDynamix?

The information technology revolution is not coming, it’s here! The winners will be shareholders in high tech stocks, college grads with science degrees and Businesses that adapt to 21st century technology.


Posted in Uncategorized | Leave a comment

Who Knew?


Figure 1


Who knew last November that the market would crash in February? That was the beginning of the Trump rally which just collapsed (Figure 1). In June of last year Kal Salama and I posted an article on Linkedin discussing the merits of the Upside Potential ratio as a predictive performance ratio and offering a link to download the software (Forsey-Sortino Model).

In September we called attention to the work of one of my former research assistants, Bernardo Kuan (deceased) who tested the efficacy of the PRI methodology and noted the methodology was “now warning investors to seek liquidity”(see “The Cost of Assuming Liquidity”) on this website.

I also called attention to a book by Joshua Cooper Ramo, Co-CEO of The Kissenger Institute. In 2016 Ramo wrote: The Seventh Sense: Power, Fortune, and Survival in the Age of Networks.  In it, he warned that we are living in the age of “collapse and construction.” He forecasted the rise of “Gatelands” controlled by a few companies with big data and massive computer power, “each, essential in some way, to national or economic security. Trade, finance, education, cybersafety, artificial intelligence, and military affairs will move from unconnected to connected.”  And, if you are not connected you are Ausgespielt (English translation SOL).  This theme was echoed by Professor Yuval Harari at the recent Davos opening ceremony.  True, they are not trained in economics or finance. If you want mathematical rigor, read “Scale, The Universal Laws of Growth, Innovation, Sustainability” by Geoffrey West, physicist and president of the Santa Fe Institute and read about “Finite Time Singularity” leading to “Entropy”.

Figure 2

Yes, the blue chart in both graphics above is the S&P 500. West also posited a new theory disputing the old economic theory of diminishing returns to size.  West explains why firms like Google and Amazon have increasing returns to size, which supports Ramo’s view of the future.

Why look outside the disciplines of economics and finance? We don’t have all the answers, and perhaps, “You won’t find a forecast of the future by sifting through the ashes of the past” (author unknown). The future may belong to big data driven by AI.

Posted in Uncategorized | Leave a comment

The Cost of Assuming Symmetry in a Skewed World

     Frank Sortino & Kal Salama

 In an Uncertain World

If you believe the uncertainty associated with investing in the indexes shown in Figure 1 is bell shaped, and it is not, all your metrics for performance measurement and asset allocation will be wrong. Well, the law of large numbers may be right in the long run, but we all live and operate in the often skewed world of the here and now.  This paper offers a tool for recognizing those times when there is more downside risk than upside potential and vise versa and how to manage it better by formally recognizing a different shape to uncertainty that is better suited to the task at hand.

That “better shape” is the brain child of two professors at Cambridge University, Aitchinson and Brown, who developed a form of the lognormal distribution that can use negative as well as positive returns to skew the distribution either left or right. Fortunately, you don’t need a PhD in mathematics to calculate the shapes shown in Figure 1.  All you need is the Forsey-Sortino model; a simplified version of software that Professor Emeritus in Mathematics, Hal Forsey and Professor Emeritus in Finance Frank Sortino developed at the Pension Research Institute (PRI) at San Francisco State University.  And all you need to know as inputs to the model is what you already know: the average return and standard deviation of each index, plus one more return that we call: the Desired Target Return®, or DTR®.  Given the better shape, the model can then calculate better statistics.  Better, in that one could have beat the market more often than not over the past 26 years.  Of course, the more people who use this model, the less inefficiencies one would expect to find in the stock market.  History indicates that will not be a problem.

                                                         Figure 1

     The picture on the left is the positively skewed large core equity index (LC) and the picture on the right is the negatively skewed small cap growth index (SG). If you are forced to viewing both of these indexes as symmetric, bell shaped distributions, you may well conclude that you should just buy the market index because in the long term, the market is bell shaped, and everyone knows, you can’t beat the market, right?  Maybe not.  The key forecasting element in Figure 1 is the Upside Potential ratio (UP ratio)[1], which was first mentioned in an article published with my Dutch friends’ way back in 1991.  PRI has repeatedly recommended replacing the Sortino ratio with the UP ratio, but to no avail.  We will conduct three tests to indicate the superiority of the UP ratio.

How it works:

Imagine two investment committees with two different investment objectives. The first investment committee is a foundation that needs to earn 6% annually to meet their scheduled philanthropic payouts. The second investment committee has a defined benefit plan and their Desired Target Return (DTR®) is the actuarial rate of return of 8.5%.  Both committees use the Forsey-Sortino model (F-S model) shown in Figure 1 that can calculate the UP ratio on data provided by their consultant.  Beginning in 1990 they both use the average return and standard deviation for the last three years as input to the F-S model.  Figure 2 shows how the foundation would key in the historic data and DTR of 6%.

                                                     Figure 2

The output from the F-S model is shown in the right hand side of Figure 2.  The UP ratio of 3.59 means this large value index they own has 3.59 times more upside potential than downside risk. The committee intuitively understands that upside potential is good and downside risk is bad.[2] They don’t know how these terms are calculated, nor do they care.  They decide to see if some combination of the nine style indexes they own could beat the market the following year by using the UP ratio.  The pension committee acted the same as the foundation, only they used the actuarial return of 8.5% as their DTR. Based on this simple framework, both committees performed three tests:

First Test: Can an Asset Allocation based on the UP ratio perform better than the market mix of 60% S&P 500 and 40% Barclays A+ Aggregate bond index? Any negative UP ratios will be allocated to the bond index.

Second test: Holding the 60/40 asset allocation constant, can the UP ratio weights select style indexes that beat the 60/40 market mix?

Third test: Can the UP ratio select equity style Indexes, remaining 100% in equities, that beat the S&P 500?

Figure 3 presents the results they would have seen on the first test year after year.

                                                            Figure 3

In 15 of the 26 years the UP ratio 6% strategy (UPR) beat the market mix. In one year it beat the market mix by less than 100 basis points so they called that a tie.  The market mix beat the UPR 6% strategy 10 times; so, using the UP ratio won  50% more often than the 60/40 market mix.  That is astounding. The pension committee, using a DTR of 8.5%, found the UP strategy beat the market mix as often as it lost but earned a higher cumulative return.  Importantly, this shows the sensitivity of the UP ratio to the return the user needs to earn in order to accomplish their investment objective.  When the committees held the asset allocation mix constant to test the ability of the UP ratio to add value, by identifying sectors of the market that would perform better than the S&P 500, they got the second test results shown in Figure 4. 

                                                                Figure 4

 Once again both UP ratio strategies beat the market mix even when the asset allocations were held to a 60/40 mix. So, it wasn’t just the asset allocation decision that was adding value. It was something buried in the calculus used to generate the picture of uncertainty shown in the F-S model output.

Third test: Can the UP ratio weighted equity style Indexes beat the S&P 500?

The all equity 6% UP ratio strategy averaged 9.63% versus 7.64% for the S&P 500 over 26 years. The all equity 8.5% UP ratio averaged 9.10% over 26 years.

The S&P 500 was only able to beat the UP ratio strategies 6 out of the 26 years and two of those times it was by less than 100 basis pts.

                                                         Figure 5

Figure 5 shows using the UP ratio weights to invest in the indexes while remaining 100% in equity performed better than the S&P 500 most of the time over the past 26 years, in spite of the fact that the three year interval following the 2008 market collapse resulted in negative UP ratios for all indexes. To ensure we were measuring 100% equity in both strategies, we assumed all funds were invested in the S&P 500 for these three years.

                                                           Figure 6

The tests above were performed with no constraints on the asset allocations. The result shown in Figure 6 identifies years when excessive allocations (100% Stock or 100% Bond) were made that could be considered outside of acceptable investment policy guidelines due to lack of diversification.  The F-S model was developed as a tool to be used by financial professionals to make better portfolio decisions. The authors caution that the model should not be the decision maker.

We now return to Figure 1, repeated below, for a more detailed explanation of the otherwise unavailable information disclosed in the F-S model output.

                                                            Figure 1

On the left is the forecast for the Surz large centrix index (LC) which had the highest UP ratio of the nine Surz Style Pure indexes®.  LC is all the stocks that are neither growth nor value.  This is different than all other core style indexes which are a mish-mosh of everything (For more details see  The UP ratio of 1.66 indicates It has 66% more upside potential than downside risk.  Because the mean is positive it is anchored at -25% causing the shape of uncertainty to be positively skewed, not symmetric and not a normal or bell shape.

The small cap growth index shown on the right has a negative mean of -3.01% so we anchored it at +25% causing it to be negatively skewed with below target risk of 23.3% as opposed to 4% for the LC index. This obviously had the worst UP ratio.  While there is a 41% chance returns could be above 6%, the UP ratio of .12 indicates this index has 88% more downside risk than upside potential.  If one doesn’t believe the inputs, you can play,” what if” games by changing them.  This is unique information one should want to know.  We don’t believe any other model has this information with only these 3 inputs.  How did these projections do so far?  As of June 1st the LC index was up 8.4% and SG was up 2%.

The Traditional Picture

                                                      Figure 7

Traditional mean-standard deviation risk measures would make the all equity UP ratio strategies appear to be the most risky over the last 26 years; highest return and highest risk. Nobody would choose the S&P 500 risk-return trade-off and everybody would choose the 8.5% UPR, really? If someone measured risk relative to their DTR and most of the volatility was above the DTR?  Risk for the committee members in this study is not a bumpy ride, it is that they don’t accomplish their investment objective, causing a failure to accomplish their goal of a specified pay out.  For them, beating the market is not the goal and is not the investment objective.  For them, the traditional framework of performance measurement is misleading.  For them, performance should be measured relative to their DTR, as shown in the Journal of Performance Measurement, Summer 2016.

Summary & Conclusions:

To prove to yourself that the Forsey-Sortino model offers you valuable information not available anywhere else, download the model as shown below and follow these steps:

  1. Enter the means and standard deviations shown at the bottom of Figure 1. Then see if the upside potential ratios give you indications of which styles to emphasize and which to deemphasize. Does this provide a unique insight?
  2. Enter the year end means and standard deviations of the three best performing and worst performing assets in your portfolio and see if the information in the F-S model would have been useful, given current results.


Please direct all responses to:

Kal Salama, CFA

Chief Investment Officer

The Headlands Group, Inc.


To download the software: Control click on the highlighted link below, install the VBruntime.exe file to ensure your computer can run software written in Visual Basic. Then install and run the F-S model.

The Forsey-Sortino Model-1 software 

This dropbox link may ask you to sign in but at the bottom you can skip this step. This zip file contains two files: Forsey-Sortino.exe and a Microsoft VBRuntime.exe.  If you get an error message with F-S Sortino.exe you must install the enclosed VB runtime.exe.

[1] Sortino, F., van der Meer, R.A.H and Plantinga, A (1999) The Dutch Triangle, Journal of Portfolio Management, Fall

[2] Researchers at Groningen University in The Netherlands found that people they interviewed intuitively understood these terms without having to explain exactly what upside potential or downside risk were or how they were calculated.


Posted in Uncategorized | 2 Comments

DTR Calculator Download

DTR® Calculator

Dr. Hal Forsey, and I, are pleased to offer another model we developed that will calculate the return a Defined Contribution participant needs to earn in order to achieve a desired pay-out.  We call this the Desired Target Return® or DTR®.[1]  Using the Department of Labor actuarial tables for male or female the model calculates the DTR with only 6 inputs.

Click here to download the DTR Calculator that will point you in the right direction:


After you download the DTR-Calculator.jar file double click on the icon.

If it does not open, download either the 32 bit (i586) or 64 bit .exe file and install.

Double Click again on the.jar icon.

You will see the following screen:

Click on the “Calculate DTR” tab.  The DTR of 7.9% will appear as shown below.  Our research indicates that 60% equity or more will be required to earn around 8%.  That’s right, this does not imply a precision to 1/10th of 1%.  This is only a guide post to get you headed in the general direction.  In our research we used 4%, 6%, 8% 10% and 12% as guideposts and we found these break points to be sufficient for portfolio construction purposes.


Suppose you are not a male.  Click the New tab and then click on the F button. Now click “Calculate” and you will see that the Department of Labor actuarial tables estimate that you will probably need to earn 9% compounded until you retire because women are expected to live longer than men.


In order for a woman to reduce her risk to that of a man of 30 she could increase her contributions.

Go up to “Current Annual Contributions”, click on the “slider” to the right of 3500 and move it right to 4550 and you will see the DTR change to 8%.

But what if you are 50 years old, not 30 years old?

Click the “NEW” tab and change the age to 50 then click, “Calculate” to see that the DTR is now 20.4%

How much longer would you have to wait to be able to retire without taking more risk?

Move the slider to the right of “Projected Retirement Age” until the DTR is 8%.  No way are you going to be allowed to work until you are 80 years old.  So start adding to your retirement assets (move slider to the right), or start an IRA.

Now Comes the Hard Part:

You will need to find someone who can construct a portfolio for you that might compound at your final DTR between now and your eventual retirement. We wish you lots of luck…

Because, all of the traditional portfolio construction frameworks available to investors ignore any future payout that will be needed and they ignore the direction to the financial destination.

The most popular framework for investing is an asset management framework, e.g., the Capital Asset Pricing Model (CAPM).  The basic assumptions of this framework are that future payouts are irrelevant, just make as much money as you can, given your tolerance for risk and hope for the best.  Proponents of this framework almost always favor a heavy allocation to equities.  The other popular framework views future payouts as liabilities and therefore uses liability driven investment strategies (LDI). This framework is almost always invested heavily in fixed income securities no matter what your contribution schedule is or where interest rates are.

To illustrate the importance of the DTR:

Assume you want to send a valuable package to New York. Asset Management advocates don’t care where you are sending it from or where you want it to go.  They assume you want to send it as far as possible as fast as they can.  LDI proponents put all packages on a truck no matter how far it needs to go or when it needs to get there.  Neither one bothers to determine the direction to send it, due to their myopic concern about which vehicle to choose.

PMPT assumes you are trying to transport your portfolio from where it is to where it needs to be at some specific time in the future in order for you to use its contents for a specified cash outflow.  Well, we built an optimizer that measured risk and reward relative to a DTR and gave that software away in a book by Dr. Steve Satchell at Cambridge University and myself.  The title is, “Managing Downside Risk in Financial Markets,”

Oh, I almost forgot.  There is one other framework that claims: all they need to know is your age In order to determine what vehicle to put your valuable package on.  It is called: “Target Date Funds” Target Date Funds also neglect to determine what direction to take.  So, they know when it has to get there, they just don’t know where or what the target is.  Low tech snake oil.




[1] Desired Target Return and DTR are registered Trademarks of Dr. Frank A. Sortino


Posted in Uncategorized | Leave a comment