Building A Simple Trading System To Diversify Your Portfolio: Part 1

Includes: SPY
by: Maik Schwäbe

Allocation within a portfolio aims to fulfill a very important goal: Stability. By diversifying between different stocks, commodities, bonds, cash and maybe even real estate, an investor seeks to break down correlations to smooth out his equity (or net wealth), usually sacrificing some of his total returns. I would like to make the case for an overlooked style of investing that every portfolio can benefit from and that I would like to think of as a separate asset class: system trading.

The definition of "system" trading spans over a whole range of different approaches but boils down to the essence of a certain set of rules that define all aspects of trading from the underlying, position size, entry and exit and leave no room for discretionary intervention.
While many investors apply certain rules or screening methods to support their investment decisions (e.g. fundamental health of a business when picking stocks for dividend investing), only a minority develops their set of rules to such a degree, that every aspect of the investment process is covered.
System-trading requires a high degree of discipline and patience and can be pretty boring, but that is true for the majority of successful investing styles.

As a trade-off there are several factors that I think every investor can benefit from by applying a 100% rule-based system to their portfolio:

  • Stability/Hedging: A good short to medium term trend following system (that allows short positions and has a routine for slow sideways markets) can be a very useful hedge to an equity portfolio, as those systems tend to have their strongest periods in volatile markets and downturns while stocks suffer the most.

  • Accountability: Of course a system has to be evaluated by a wide range of performance and risk related metrics prior to being traded with real money. Thanks to the flood of free data and information and ever advancing software available, it has become pretty easy for every individual to backtest an systematic approach. Therefore it is clearly visible in what kind of market-phases systems perform and fail and how much pain an investor has to be prepared to take.
    Without a clear track record, the same can not be said about discretionary investment decisions, or as I like to put it: You can't backtest your guts!

  • Peace of mind: I don't take a lot of discretionary trades but when I do, they are always accompanied by second thoughts; for example: I own a stock for the long run because I am convinced of its fundamental prospects for the future, but there was a huge and fast run up which leads me to think of taking profits prematurely.
    Those disturbing thoughts never pop up while trading a rule based system where every entry and exit is 100% defined. So system trading can be especially comfortable for less confident investors who tend to doubt their own decisions.

Goals and Inspiration

There are a thousand different ways and approaches that can lead to success and in this and some upcoming articles I want to show one. The goal is to create a 100% rule based system that is easy to trade and doesn´t need a lot of attention or maintenance. A stable equity curve gets more priority than absolute performance, and the turnover should be relatively low to avoid a negative impact due to transaction costs.
It would also be very pleasant if the system could show resilience in times of market turmoil, so that it could be seen as a stabilizing addition to an equity portfolio.
I will try to fit several seasonal and event based trades with favorable odds into one system to reach the above mentioned goals. A lot of inspiration for that approach came from Rob Hannas' Quantifiable Edges Blog.
This system will still leave a lot of room for own interpretation and additions, and even if this approach does not appeal to an individual investor, there are still a lot of interesting takeaways for everybody in terms of seasonalities.

Part 1: Fed-Days

It is a well reported fact that the days on which the Fed announces their policy and rate changes have a bullish bias. The logic behind it is clear: The Federal Reserve is one of the most powerful institutions to influence the real economy and thus the financial markets and their decisions are usually aimed to promote or support economic stability.

But just how strong is that bullish bias? Is it really an edge or just coincidence? And can it be exploited in a sustainable way?

So let us define our criteria for this test:

  • We buy the S&P 500 on the close of the day before the Fed announces its policies, hold the position for exactly one day and sell on the close of the day of the Fed announcement.

  • We only consider the scheduled Fed meetings (eight per year)

  • Our test period will be from March 1st 1988 until March 1st 2013, so exactly 25 years.

  • No stop orders, slippage or transaction fees will be considered

The S&P 500 cash data is easy to obtain, and the question of which instrument to trade the strategy with (Futures, (NYSEARCA:SPY) etc.) will be discussed at the end of this series of articles.

System S&P 500
Trades total: 200 Days total: 6297
Winning Trades: 118 (59%) Up Days: 3372 (53,5%)
Losing Trades: 82 (41%) Down Days: 2925 (46,5%)
Avg. Winner: +0,959% Avg. Profit: +0,758%
Avg. Loser: -0,632% Avg. Loss: -0,8%
Max. Winner: +5,14% Max. Profit: +11,58%
Max. Loser: -2,94% Max. Loss: -9,03%
Expectancy*: +0,306% Expectancy:


Performance: +82,2%

If we compare those statistics to the total of trading days in the S&P500 in those 25 years, the presented upside edge becomes visible in the increased winning percentage and especially in the ratio between average winners and losers.

Now that we have a proven bullish bias on Fed days, let's take it one step further.
If the Fed has such a positive influence on the markets performance, one would suggest that this would be especially true in times of market turmoil. So let's add a simple filter to split our results between stable and unsettled market conditions. How does our strategy perform when the days of our entries close above/below the 200 day moving average?

All other parameters are the same as in the test above.

If Close above 200 SMA
System S&P 500
Trades total: 146 Days total: 4511
Winning Trades: 88 (60,3%) Up Days: 2435 (54%)
Losing Trades: 58 (39,7%) Down Days: 2076 (46%)
Avg. Winner: +0,759% Avg. Profit: +0,606%
Avg. Loser: -0,515% Avg. Loss: -0,632%
Max. Winner: +2,92% Max. Profit: +5,12%
Max. Loser: -2,53% Max. Loss: -6,87%
Expectancy: +0,253% Expectancy: +0,036%
Performance: +43,8%
If Close below 200 SMA
System S&P 500
Trades total: 54 Days total: 1786
Winning Trades: 30 (55,6%) Up Days: 937 (52,5%)
Losing Trades: 24 (44,4%) Down Days: 849 (47,5%)
Avg. Winner: +1,545% Avg. Profit: +1,153%
Avg. Loser: -0,912% Avg. Loss: -1,211%
Max. Winner: +5,14% Max. Profit: +11,58%
Max. Loser: -2,94% Max. Loss: -9,04%
Expectancy: +0,454% Expectancy: +0,03%
Performance: +26,7%

As you can see there is no huge difference between the two setups. While the winning percentage in unsettled conditions decreases slightly, all other metrics leap up, pointing to a generally higher volatility and not necessarily a higher probability of an upside day.

Unscheduled Meetings

Over the years the Fed has held a number of unscheduled meetings, usually in the aftermath of especially turbulent events such as Lehman in 2008 and the Flash Crash. Due to the fact that those meetings might happen spontaneously and sometimes even on Sundays, it is hard to measure the market's performance on those days.
As far as I could see, there were 10 unscheduled meetings on trading days, all of them since January 2008. Four of those turned out to be decent up-days, another 4 ended modestly in the red while two days during October 2008 ended with -5% and nearly -9%.
Having only 10 data entries with such skewed results and considering the difficulty of reacting to quickly announced meetings that can happen over the weekend, I decided to not add these events to the original setup.


Even if a system provides a 100% hit rate, it would not be complete without protection to prevent it from destroying a year long edge with a few single catastrophic losses.
In our case, the stop loss has to give the trade some air to breathe but should make a clear cut when things turn ugly. One could apply a volatility or standard deviation based stop which all are valid alternatives, but to keep things very simple I will use a fix stop 1.5% away from the entry.


  • Buy the S&P500 on the close of the day before the Fed announces its policies
  • Sell at -1.5%, or the close of the next trading day

  • Only scheduled Fed meetings (eight per year)

  • March 1st 1988 to March 1st 2013

  • No slippage or transaction fees will be considered

Final Results
Trades total: 200
Winning Trades: 114 (57%)
Losing Trades: 86 (43%)
Avg. Winner: +0,907%
Avg. Loser: -0,63%
Max. Winner: +5,14%
Max. Loser: -1,5%
Expectancy: +0,246%
Performance: +61,7%

System Equity

As expected the number of positive trades declines a little (by four), as we get stopped out before the market actually reversed higher and ended in positive territory (which actually speaks for the theory of the Fed as a supporting role in the markets). The declining avg. winner and overall performance indicates that those four trades were contributing quite a bit to the bottom line, but that is the price one has to pay for the peace of mind of never running unprotected into a catastrophic plunge.


We did confirm a positive edge and built a very simple 100% rule based strategy around it. With eight days a year of exposure it can hardly be called a complete system, but it is the first piece of a bigger puzzle.

In upcoming articles I will introduce additional pieces to make this system more active, and also a routine that evaluates if a certain edge still makes sense trading or should be abandoned at all. Because even if something is working for a quarter of a century it doesn´t mean that it can´t break down completely.

Disclosure: I am long SPY. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

About this article:

Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500.
Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here