Questions & Answers Forum

Direct any general questions you may have on the topics of Finance, Financial Markets, Investments, Trading, etc.

Diversification Basics

               I have heard many different opinions on what constitutes proper diversification.  First, we can ask the question, “What is diversification?”  Simply put, the answer is; not putting all your eggs in one basket. It is a technique that helps minimize risk by spreading your assets across different types of investments.  So to the extreme, it is putting each of your eggs into separate baskets.

               Most risk levels (see the Various Types of Risk in the Market post on this blog) can be diversified out.  Ideally, you would want to spread your assets across as wide a range as possible.  Currently, higher costs and complications (int’l taxes, currency exchange fees, etc.) make some of this unfeasible.  It would be hard for the average American who was managing their own portfolio to buy shares of a tire company traded on the Australian exchange, a tea producer on the Hong Kong exchange, a bank traded on the London exchange, etc.  Fortunately, the vast expansion of the internet does allow that person the ability to diversify quite adequately.  Here are some of the arguments and their underlying reasoning:

·         Most finance majors are taught in college that if the stock-price section of the Wall Street Journal is hung on the wall and a blindfolded individual throws darts until ten stocks are hit, that a portfolio (or a basket) of those investments is purchased will perform almost identically to any market index.  Broken down, any ten randomly selected investments will perform up to par with the market.
 

·         Dr. Jim Cramer (host of Mad Money on CNBC) holds a philosophy that a portfolio of five strategically selected stocks will be adequately diversified.  The selected stocks should be chosen from five different sectors/industries and that an evaluation of the fundamentals (or key strength indicators) should be done to select firms that are one of the best-of-breed in their respective industry.  The result is proper diversification and a little additional safety from some firm specific rick elements (bankruptcy, etc.).
 

·         Statistics teaches that an accurate representation of a group (the markets, in this case) can be reach with a random sample (portfolio) size of 30.  Interpreted, a portfolio of 30 randomly selected investments will perform 99.9% similarly to the market.  Our most common market indexes fall into this general category (while ignoring the benefit of random selection).  The S&P500 is a summation of the performance of 500 different stocks.  The Dow Jones Industrial Average is a function of 30 notable industrial stocks.  It is an important note that while the DJIA is an indicator of a broad industrial sector, it also is used to represent the overall health of our domestically traded market.
 

My philosophy is based upon these teachings, my trading style, and the experience I have trading in the market.  I tend to take a more statistical approach when selecting stocks to create a diversified portfolio from.  I generally select 20 stocks for my trading basket.  Why not 30?  While statistical concepts are very well founded, when applied to portfolio selection – the added transaction costs outweigh the benefit of adding the 10 additional stocks.  It costs [me] roughly $10 to buy, and then another $10 to sell a stock in the market.  The benefits of additional diversification from adding the 21^st stock to the basket are outweighed by the transaction fees.  As a fund gets larger, it may indeed become more cost effective to add the additional stock(s), but in reality the effect is minimal.
 

Often, my model will output two stocks within the same sector (example: Coca-Cola and Pepsi).  I generally do not have to be concerned with this happening because I have a basket of 20, not 5.  I will express that each additional stock within the same sector as another exposes the portfolio to more industry/sector risk.  If I get more than three stocks from the same industry, I will dump one out for the next stock suggested by the model.
 

Why not use the random-dart-toss method?  Let’s face it – Any person educated in finance should be able to select better picks than what would result.  Cramer’s “Best-of-Breed” teachings will even help produce a comparably diversified portfolio that will likely produce higher gains than a completely randomly selected portfolio would.  Other factors also make random selection foolish, such as cyclical stocks (to be discussed in another Blog post), current trends or firm specific factors that would raise a red flag in the minds of even the average investor.
 

The lesson to take from all of this is that most of the risk of loss can be minimized for any investor through simple and proper diversification.  However; diversification will not reduce [global] market risk (which can be reduced; See my blog topic on Hedging *Not Yet Posted*).

Various Types of Risk in the Market

                For the purposes of this discussion (and most in Finance), one should think of the term risk as an uncertainty, specifically as downside risk or the risk of loss.  No trader or investor worries about the uncertainty that they may unexpectedly see higher gains in their investments.  So we will ignore the upside for the purposes of this topic.  Risk can be divided into a type by the level of the market that it affects.  We will go over the broad types:

• Firm-Specific Risk: The risk that a particular company will perform poorly, or worse – fail altogether.  For example, company ABC releases their quarterly report and falls short of expected earnings for whatever reasons (increased cost of products/services, decrease in demand for their products/services).
• Sector/Industry Risk: The risk that a particular industry or sector will perform poorly (most/all firms within that sector).  The development of file-sharing impacted the music industry significantly, and had a negative impact on the entire sector – from the writers all the way through retail music stores.
• [National] Market Risk: The risk that an entire market will degenerate (most/all sectors within the market).  A simple example of this, and probably one that most of you will remember, was the .com bubble bursting at the turn of the millennium.  It negatively affected the entire market system here in the US.
• Global Market Risk*:  This is solely in my opinion, but I feel that the world market as a whole has a large enough of a degree of separation from individual [national] markets for a distinction to be made in the terms of risk.  The recent Real Estate bubble proliferated through and pulled down not only our national market here in the US, but had major negative effects on the world market.  It encompassed most of North America and Europe, and it impacted markets everywhere in the world.

Many times, firm specific risk is at the root of the other levels of risk. The housing bubble (2008) began in the financial sector, with mortgage lenders relaxing important standards for lending.   It more than likely started with one firm easing its standards in order to gain more business over competitors.  Competitors soon then followed suit in order to balance the odds.  Had the repercussions of lax standards been realized much sooner, the impacts would not have filtered up through every market in the world.
 

Other events however, can bypass firm and even sector risk.  A severe drought, for example would hurt our economy overall – beginning in the commodities markets (corn, cotton, beef, wheat, other produce) and affect many sectors at once.

What does all this mean? Virtually all risk can be largely reduced. Two common practices for avoiding risk are:

Diversification: Spreading assets across different investment types.  See my blog topic on Diversification Basics.
Hedging: Placing your assets on ‘both sides of the fence’.  See my blog topic on Hedging*Not Yet Posted*

* In general, the term ‘Market Risk’ includes the world market as well.  However; I argue that there should be a distinction between this and the domestic market.

Our History

                The foundations of our research efforts began in 2006.  I began trading in 2004 after a friend of mine turned $5,000 into over $30,000 in a few short months, on a penny stock.  In hindsight, he was very lucky to have realized that outcome.  Over the next few years I traded, in what I would now admit was mostly ignorance.  I would pick a company and put money into the market with little more than a 50/50 shot of making anything.

                It was also in 2004 that I started my Finance degree, after more than ten years of working in the factory.  As my education advanced, I began to develop insights into how vastly public information could be used to make more informed decisions on which companies held a greater chance of growing my investment dollar.  The financial discipline was not the only knowledge that would help me to invest.  I realized that Statistics could also play a large role (and arguably a more important one) with respect to my investment decisions.

                In December of 2006, I began developing a statistical model for evaluating large numbers of companies at once.  The results of the model are then filtered through an additional algorithm (is a step-by-step procedure for calculations) to produce a group of potential investments which have a lower level of risk as compared to other choices.  The initial trials went very well in 2007 and the beginning of 2008.  The model stopped producing results when the markets began to destabilize in mid-2008, which it was intended to do.

                Unfortunately because of the market crash, the program would simply yield no results.  That fact halted any further research and the project was archived.  After completing my Master’s Degree, I began revisiting the idea of continued research.   Last fall (2012), I began running multiple models again and sharing the stocks that the model produced, and then following the resulting gains/losses and benchmarking those results against the S&P 500.  The most impressive model produced a gain of 10.52% over a 9 week period, and the least impressive model resulted in a gain of only 2.51% in the same time period.

                Several weeks ago, my colleagues and I discussed making this research more public.  As this has become a hobby of mine, I chose to create this blog to discuss my research, document the portfolios of each model and the future results of the trials.  Additionally, I will be happy to discuss and teach various topics within the discipline of Finance with my readers.

                We will be running the first set of models (for this blog) on Sunday, March 3^rd, and posting the resulting theoretical portfolios.