Investing | Nitin Gupta

Investment Strategy Diversification

Tue, 27 Jun 2017 00:00:00 +0000

By all conventional and even some unconventional measures, the US stock market is trading way beyond historical valuation averages and is closer to all time highs. Passive stock index investors have enjoyed a period of extraordinary gains in one of the longest running bull markets. No other major asset class has come close in the last 7 years. Well diversified portfolios have had lackluster returns while the stock market keeps making new highs.

	US Stocks	Foreign Stocks	US Bonds	Foreign Bonds	Treasuries	Gold	Commodities	REITs
Table 1: Performance by Asset Class
7 Years
May 2010 - Apr 2017 Annualized Return	12.72%	4.47%	3.36%	0.79%	7.28%	0.66%	-6.91%	11.09%
5 Years
May 2012 - Apr 2017 Annualized Return	13.54%	5.40%	2.23%	-0.61%	3.53%	-5.69%	-12.23%	9.16%
3 Years
May 2014 - Apr 2017 Annualized Return	10.35%	1.09%	2.64%	-3.14%	5.95%	-0.93%	-17.55%	8.82%

But overstretched valuations in a contentious political climate raise fears of what’s around the corner. Small market gains accumulated gradually over months could be wiped out within days or weeks, when the sentiment turns. So what should investors do?

Momentum strategies provide an attractive alternative in this scenario. Momentum has been described as a premier market anomaly. Studies have been published for over 2 decades showing why and how well it works. Despite being well publicized in the investment world, momentum strategies have continued to work well over long periods of time.

Some of the biggest advantages of momentum strategies are:

Their entry and exit rules could be well-defined and understood
They could be implemented using highly liquid, low cost ETFs
They are relatively easier to scale and execute
During market downturns, these strategies exhibit lower volatilities and drawdowns as compared to value and passive investing strategies

That being said, momentum is not a single monolithic strategy as many academics portray it to be. It could be implemented in many different ways. Here I present three momentum strategies that could be combined in simple ways to mitigate overall volatility and drawdowns, while maintaining attractive returns.

Global Equities Momentum Strategy

The first of these is Global Equities Momentum (“GEM”), which has been detailed by Gary Antonacci in his fantastic book, Dual Momentum Investing. It is one of the simplest strategies to understand and implement. GEM effectively makes use of zero correlation betweeen stocks and bonds. It utilizes absolute momentum to switch to bonds when stocks show weak performance, and bonds tend to do well. Additionally, it uses relative momentum to switch between domestic and foreign stocks during times of stock market strength.

Here are the concrete rules, dataset and assumptions to backtest GEM:

Rules : Between domestic and foreign equities, invest in the one which had higher outperformance relative to the risk free rate in the past 12 months. Otherwise invest in bonds.

Dataset : Total returns calculated from daily adjusted closing prices downloaded from Yahoo! Finance

Domestic stocks (SPY, combined with VFINX prior to Feb 1993)
Foreign stocks (CWI, combined with VGTSX prior to Jan 2007)
Bonds (AGG, combined with VBMFX prior to Oct 2003)
Risk free rate determined by T-bills data downloaded from FRED.

Assumptions : Signals and weights determined at the end of month (t), rebalanced at the end of first business day of next month (t+1), transaction costs 20 bps per trade.

	SP500TR	GEM_Antonacci	GEM	EqualWts
Table 2: Comparison of GEM strategy with an equally weighted portfolio of SPY, CWI and AGG
May 1997 - Apr 2017 Annualized Return	7.52%	12.66%	11.79%	6.24%
May 1997 - Apr 2017 Annualized Std Dev	15.17%	11.93%	11.97%	10.68%
May 1997 - Apr 2017 Annualized Sharpe (Rf=2.04%)	0.35	0.87	0.80	0.38
May 1997 - Apr 2017 Worst Drawdown	-50.95%	-17.84%	-19.32%	-38.77%

It is easy to cross-check the backtesting results with simulated results from Gary’s website. I do not expect to match the annualized returns as Gary’s simulations use theoretical indexes, do not apply any fees or transaction costs, and monthly performance numbers on his website are rounded to the first decimal place. But a correlation of monthly returns of 0.99, annualized standard deviations, shape of the equity curves and the close alignment of drawdowns all suggest a close match.

The GEM strategy performance shows simple rules could be used for timing entries and exits and work exceedingly well over complete market cycles. GEM trounces the S&P 500 Total Returns Index on both absolute and risk adjusted basis over 2 market cycles in this backtesting period.

The results of an equally weighted long only strategy of SPY, CWI and AGG are also shown. Essentially it’s a simplified and slightly aggressive 66.67% stocks and 33.33% bonds portfolio, which gives a sense of the performance of a passive structurally allocated portfolio strategy. As compared to the broader market, it does reduce the volatility but a worst drawdown of ~ 40% is still driven by stocks.

The edge provided by simple market timing rules in GEM is evident as compared to an equally weighted long only portfolio of the 3 underlying assets. The drawdowns are much lower as there’s a clear rule for exiting the market when it underperforms relative to the risk free rate. Moreover, there have been very few trades over the past 20 years, and most of the gains have been long term in nature. Hence GEM is quite attractive from a tax standpoint too.

SPY-TLT Universal Investment Strategy

This strategy has been detailed on LogicalInvest. It utilizes just 2 asset classes, stocks and long term treasuries, and takes advantage of a slighly inverse correlation between the two. While GEM rotates between stocks and bonds in a 100% risk-on / risk-off manner, this strategy is more nuanced. It rotates between stocks and long term treasuries in a more measured way. It selects between a mix of stocks and treasuries which has performed the best on a risk-adjusted basis, in the recent past.

Here are the concrete rules, dataset and assumptions to backtest UIS:

Rules : Construct 11 portfolios ranging from 100% SPY, 0% TLT to 0% SPY, 100% TLT, by successively incrementing the weight of TLT by 10%. Determine the returns and volatility of these 11 portfolios in the past 60 days (~ 3 months). Select the portfolio that gives the best risk adjusted returns in this period and apply the same weights to the following month.

Dataset : Total returns calculated from daily adjusted closing prices downloaded from Yahoo! Finance

Stocks (SPY, combined with VFINX prior to Feb 1993)
Long-term Treasuries (TLT, combined with VUSTX prior to Aug 2002).

Assumptions : Weights determined at the end of month (t), rebalanced at the end of first business day of next month (t+1), transaction costs 20 bps per trade.

	SP500TR	UIS	EqualWts
Table 3: Comparison of UIS strategy with an equally weighted portfolio of SPY and TLT
Sep 1986 - Apr 2017 Annualized Return	10.01%	11.46%	9.12%
Sep 1986 - Apr 2017 Annualized Std Dev	14.99%	11.22%	8.83%
Sep 1986 - Apr 2017 Annualized Sharpe (Rf=3.23%)	0.44	0.71	0.64
Sep 1986 - Apr 2017 Worst Drawdown	-50.95%	-23.51%	-22.35%

The performance of both UIS and an equally weighted long only mix of stocks and treasuries show the benefits of diversication between these 2 asset classes. In fact an equally weighted portfolio would’ve done just about as well as the S&P 500 Total Returns Index with far lower volatility and drawdowns. Since the global financial crisis of 2008, UIS has performed exceedingly well, which has led to its outperformance on both absolute as well as risk-adjusted basis.

Three-Way Momentum Strategy

This strategy is a more simplified version of a three-way momentum model posted by Meb Faber on his blog, that was originally done by Ned Davis Research. As Meb has suggested, it is conceptually similar to the strategies in his QTAA paper. This strategy is also one of the simplest to understand and implement. It utilizes just 3 asset classes: stocks, bonds and gold and invests equally in all asset classes that show positive momentum.

Here are the concrete rules, dataset and assumptions to backtest this strategy:

Rules : Invest equally in stocks, bonds and gold as long as they trade above their past 200 days moving average. If none of them trades above their 200 days moving average, then invest in T-bills.

Dataset : Total returns calculated from daily adjusted closing prices downloaded from Yahoo! Finance

Stocks (SPY, combined with VFINX prior to Feb 1993)
Bonds (AGG, combined with VBMFX prior to Oct 2003)
Gold (GLD, combined with LBMA/GOLD prior to Nov 2004).
Risk free rate determined by T-bills data downloaded from FRED.

Assumptions : Signals and weights determined at the end of month (t), rebalanced at the end of first business day of next month (t+1), transaction costs 20 bps per trade.

	SP500TR	ThreeWay	EqualWts
Table 4: Comparison of Three-Way Momentum Strategy with an equally weighted portfolio of SPY, AGG and GLD
Nov 1987 - Apr 2017 Annualized Return	10.31%	8.60%	7.05%
Nov 1987 - Apr 2017 Annualized Std Dev	14.31%	7.57%	7.22%
Nov 1987 - Apr 2017 Annualized Sharpe (Rf=3.13%)	0.48	0.70	0.52
Nov 1987 - Apr 2017 Worst Drawdown	-50.95%	-13.34%	-18.71%

Strategy Correlations

Although GEM works very well on its own, it is unlikely that the other two strategies could be followed on a stand-alone basis. Even the 100% risk-on / risk-off nature of GEM make it difficult to employ it at scale.

But let’s look at the correlations of these strategies:

	SP500TR	GEM	UIS	ThreeWay
Table 5: Correlation of monthly returns (May 1997 to April 2017)
SP500TR	1.00	0.67	0.26	0.35
GEM	0.67	1.00	0.31	0.55
UIS	0.26	0.31	1.00	0.37
ThreeWay	0.35	0.55	0.37	1.00

The UIS strategy has low correlations with both GEM and the Three-Way strategy. Even GEM and the Three-Way Strategy have a correlation of ~ 0.55. So they could be effectively blended together to form a portfolio of diversified strategies.

Here are a couple of simple ways to blend these strategies:

Strategy Blends

In the last 20 years backtesting period, an equal weight combination of GEM and UIS results in a significant reduction in portfolio volatility without compromising returns. Hence the blended strategy gets a solid boost on risk adjusted returns.

	SP500TR	GEM	UIS	Equal_Blend
Table 6: Comparison of a blended strategy with 50% each of GEM & UIS strategies
May 1997 - Apr 2017 Annualized Return	7.52%	11.79%	12.10%	12.20%
May 1997 - Apr 2017 Annualized Std Dev	15.17%	11.97%	11.15%	9.38%
May 1997 - Apr 2017 Annualized Sharpe (Rf=2.04%)	0.35	0.80	0.88	1.06
May 1997 - Apr 2017 Worst Drawdown	-50.95%	-19.32%	-17.81%	-11.81%

Combining GEM, UIS and Three-Way strategies on an equal weight basis, leads to even further reduction in portfolio volatility. At 8% it is closer to the lowest volatility of the three strategies. Even though the annualized return of the blended strategy is slightly lower than that of GEM, the risk adjusted return is ~ 40% higher.

	SP500TR	GEM	UIS	ThreeWay	Equal_Blend
Table 7: Comparison of a blended strategy with 33.33% each of GEM, UIS & ThreeWay strategies
ALL Years
May 1997 - Apr 2017 Annualized Return	7.52%	11.79%	12.10%	9.34%	11.32%
May 1997 - Apr 2017 Annualized Std Dev	15.17%	11.97%	11.15%	7.72%	8.00%
May 1997 - Apr 2017 Annualized Sharpe (Rf=2.04%)	0.35	0.80	0.88	0.93	1.14
May 1997 - Apr 2017 Worst Drawdown	-50.95%	-19.32%	-17.81%	-9.39%	-9.46%
10 Years
May 2007 - Apr 2017 Annualized Return	7.00%	6.86%	16.07%	8.39%	10.69%
May 2007 - Apr 2017 Annualized Std Dev	15.25%	11.83%	12.56%	8.13%	7.91%
May 2007 - Apr 2017 Annualized Sharpe (Rf=0.52%)	0.42	0.53	1.23	0.96	1.28
May 2007 - Apr 2017 Worst Drawdown	-50.95%	-19.32%	-14.40%	-9.39%	-9.46%
5 Years
May 2012 - Apr 2017 Annualized Return	13.36%	9.05%	10.24%	5.29%	8.25%
May 2012 - Apr 2017 Annualized Std Dev	10.19%	9.00%	8.08%	6.70%	7.07%
May 2012 - Apr 2017 Annualized Sharpe (Rf=0.15%)	1.29	0.99	1.25	0.77	1.14
May 2012 - Apr 2017 Worst Drawdown	-8.36%	-14.52%	-10.50%	-5.04%	-9.46%
1 Year
May 2016 - Apr 2017 Annualized Return	17.77%	17.77%	17.53%	8.05%	14.43%
May 2016 - Apr 2017 Annualized Std Dev	6.05%	6.05%	6.34%	7.32%	5.52%
May 2016 - Apr 2017 Annualized Sharpe (Rf=0.44%)	2.85	2.85	2.69	1.04	2.52
May 2016 - Apr 2017 Worst Drawdown	-1.73%	-1.73%	-3.02%	-5.04%	-2.44%

Here are the equity curves and the distribution of weights in the blended strategy:

Advantages

There are many advantages of this blended strategy:

Works well in both bull and bear markets with low volatility
Uses low cost, highly liquid ETFs
With just 5 ETFs, it is very easy to implement and execute
Low cost substitute ETFs and funds exist to employ it at scale
Highly competitive with more complex strategies

Leveraged Blends

Given that the blended strategy exhibits very low volatility, the returns could be boosted by using leverage. To get a good sense of leveraged returns, I benchmark the margin rate at the prevailing 1-month USD LIBOR rate + 2%. This is well above the rates offered by brokerages catering to professional investors.

At the same level of volatility, the blended strategy would have produced over twice the returns of S&P 500 Total Returns Index, with much lower drawdowns.

	SP500TR	Blend_1X	Blend_1.3X	Blend_1.5X	Blend_2X
Table 8: Comparison of leveraged blends
ALL Years
May 1997 - Apr 2017 Annualized Return	7.52%	11.32%	13.30%	14.60%	17.80%
May 1997 - Apr 2017 Annualized Std Dev	15.17%	8.00%	10.40%	11.99%	15.99%
May 1997 - Apr 2017 Annualized Sharpe (Rf=2.04%)	0.35	1.14	1.06	1.03	0.97
May 1997 - Apr 2017 Worst Drawdown	-50.95%	-9.46%	-12.48%	-14.46%	-19.30%
10 Years
May 2007 - Apr 2017 Annualized Return	7.00%	10.69%	12.99%	14.52%	18.30%
May 2007 - Apr 2017 Annualized Std Dev	15.25%	7.91%	10.29%	11.88%	15.86%
May 2007 - Apr 2017 Annualized Sharpe (Rf=0.52%)	0.42	1.28	1.21	1.17	1.12
May 2007 - Apr 2017 Worst Drawdown	-50.95%	-9.46%	-12.48%	-14.46%	-19.30%
5 Years
May 2012 - Apr 2017 Annualized Return	13.36%	8.25%	9.98%	11.13%	13.93%
May 2012 - Apr 2017 Annualized Std Dev	10.19%	7.07%	9.20%	10.61%	14.14%
May 2012 - Apr 2017 Annualized Sharpe (Rf=0.15%)	1.29	1.14	1.07	1.03	0.97
May 2012 - Apr 2017 Worst Drawdown	-8.36%	-9.46%	-12.48%	-14.46%	-19.30%
1 Year
May 2016 - Apr 2017 Annualized Return	17.77%	14.43%	18.13%	20.65%	27.08%
May 2016 - Apr 2017 Annualized Std Dev	6.05%	5.52%	7.18%	8.28%	11.04%
May 2016 - Apr 2017 Annualized Sharpe (Rf=0.44%)	2.85	2.52	2.46	2.43	2.40
May 2016 - Apr 2017 Worst Drawdown	-1.73%	-2.44%	-3.35%	-3.96%	-5.47%

Further Ideas

Scaling: Create weekly tranches, check signals and rebalance weekly instead of monthly. This way only 25% of the portfolio is exposed to whipsaw risk. Also, the trade sizes are reduced at any given point. I wrote about these ideas in a prior post.
Risk Management: From the weights plot of the blended strategy, it is evident that during some periods, 100% of the portfolio risk is concentrated in SPY. A prudent risk management strategy would limit the concentration on any single asset in a portfolio. So sudden unfavorable movements have less of an impact.
Add more uncorrelated strategies: Instead of being all-in on momentum only, value strategies could be added to the mix, which exhibit low correlation to momentum strategies.
Alternative Implementation: Implement this strategy using futures products which provide greater control over margins to scale up or down.

Conclusion

Investors are either totally passive or too fixated on finding a single unicorn strategy that’ll work well during both good and bad times. The key is to put together a team of sufficiently uncorrelated strategies that work well together to build a robust portfolio.

Disclaimer

All content displayed here is for informational purposes only and is not guaranteed to be accurate, complete or up-to-date. The returns presented are hypothetical and do not represent returns attained by any investor. The content and commentary are intended to provide the views and observations of the author only, and are subject to change at any time without prior notice, and do NOT represent those of past, present or future employers. Nothing herein should be considered investment advice or recommendation to buy, sell or hold any securities.

Luck in Rebalance Timing

Mon, 29 May 2017 00:00:00 +0000

An important and often overlooked topic was raised by Corey Hoffstein at NewFound Research. Here are his first couple of tweets on that topic:

Indeed, this is about rebalance timing and how little attention it gets. Within the construct of a systematic strategy, this is a part of the Execution Model.

The Black Box Revealed

Courtesy: Inside the Black Box - The Simple Truth about Quantitative Trading by Rishi Narang

Unless it’s an intraday trading model, when it comes to portfolio rebalancing, a common convention is to rebalance at the beginning of a new month. Perhaps, these are some of the reasons behind this convention:

In published literature, a monthly period is commonly used for portfolio rebalancing. Hence it has become a sort of common convention to follow.
Most of the researchers are adept in using Excel, but less so in programming languages. While doing a backtest in Excel, it is easier to manipulate monthly data and do performance analysis, as opposed doing it with daily data.
In some cases, investible products for some asset classes have not existed beyond a couple of decades at best. So longer backtests are often done using theoretical indices which are often available only in monthly format. Hence it forces the choice of a monthly rebalancing cycle. This implies rebalancing at the beginning of a new month.

But should this be accepted just on the basis of a common convention? If a monthly rebalancing cycle is adopted, does it matter which day of the month is chosen for rebalance? Are there any major differences between portfolios rebalanced on the first business day of a month, as opposed to other days? Are these differences just completely random in nature?

These are some of the questions that need to be answered.

Strategy

To answer these questions, I used one of the most basic market timing strategies:

Go long S&P500 when it trades above its 200 days simple moving average (SMA), otherwise exit and invest in T-Bills.

There is nothing magical about using a 200 days SMA. Other periods in the same ballpark work just as well, as shown by Meb Faber in his QTAA paper. I used this simple strategy to illustrate the effect of rebalance timing on portfolio performance.

I put together a historical daily total returns dataset of SPDR S&P 500 ETF (SPY), spliced together with historical daily total returns data of Vanguard 500 Index fund (VFINX), prior to Feb. 1993. This data was sourced from Yahoo! Finance, prior to their API fiasco earlier this month. Both of these products closely track the S&P500 Total Returns Index. This provides 37+ years of historical daily total returns data going back to the beginning of 1980.

Methodology

To test the effect of rebalance timing, I constructed 21 portfolio series, assuming 21 business days in a month. The first portfolio series is rebalanced on the first business day of a month, using the trading signal as of the end of the previous month. In short, signal @ t, rebalance @ t+1. If there are fewer than 21 business days in a month, say for instance 19 days, then portfolio series 20 and 21 are rebalanced on the first business day of the next month. A transaction cost of 10 bps is applied on each trade.

Here are the results:

The impact of luck could be observed in drawdowns as well as returns. The worst drawdown of series 13 is just a little more than half of series 14. Nevertheless, their annualized returns are almost the same. In all the cases, except series 13, the worst drawdown happened after the crash of 1987. Ironically, series 13 is the lucky one here ;). As Corey mentioned in his tweets, for an investment manager this is a difference between getting hired and fired.

At the moment, series 1, gives the best return. As we progress to series 9, the returns decrease quite significantly. There is a difference of over 2% in their annualized returns. Here are their performance summaries:

	Best	Worst
Table 1: Best and worst rebalance series
ALL Years
Jan 1981 - Apr 2017 Annualized Return	10.68%	8.63%
Jan 1981 - Apr 2017 Annualized Std Dev	11.09%	11.87%
Jan 1981 - Apr 2017 Annualized Sharpe (Rf=4.22%)	0.55	0.35
Jan 1981 - Apr 2017 Worst Drawdown	-23.59%	-26.20%

Let’s take a look at how the best (blue) and the worst (red) series have fared relative to the other 19 series in the entire historical period:

So, turns out the current best performed worse than the median for most of the historical period. But since 2003, it really took off and has managed to avoid many whipsaws that appear to have affected other series. But will it continue to do better? The impact of luck is quite obvious here. Quite surprisingly, the worst series has been at or near the worst throughout the historical period. Perhaps a longer backtest or one with a different investment strategy would be useful to investigate this further.

Mitigating the impact of luck

So how could the impact of luck be mitigated without resorting to more frequent rebalancing, and thus incurring higher portfolio turnover costs?

Here are a couple of ideas:

Create four weekly portfolio series staggered by a week, where each series is rebalanced at the interval of 4 weeks. This way, at any given time, only ~ 25% of the capital is exposed to whipsaw risk. But a practical challenge here is to ensure that each series remains balanced relative to others. Say for instance, no single series should have a weight greater than a third of the entire portfolio. This could be managed by leveraging / deleveraging at rebalance time.
Stick with rebalancing at the beginning of a new month (t+1). But instead of taking the signal at the last day of the prior month (t), take the median (or mean) of the daily signals in the prior month. This is similar to Corey’s idea of signal smoothing with 21 days SMA of signals.

Let’s see how these methods compare to the best and worst portfolios:

The three methods with overlapping portfolios and signal smoothing are denoted by Weekly_avg, Signal_median, Signal_avg. We can observe that all of these methods have similar performance profiles in the historical period, with minor differences in shorter time periods. They perform much better than the Worst portfolio series, in all time periods. And during most of the historical period, they have even performed better than the current Best series.

	SP500TR	Best	Worst	Weekly_avg	Signal_median	Signal_avg
Table 2: Comparison of rebalance methods
ALL Years
Jan 1981 - Apr 2017 Annualized Return	11.07%	10.68%	8.63%	9.99%	10.19%	10.26%
Jan 1981 - Apr 2017 Annualized Std Dev	14.85%	11.09%	11.87%	11.01%	11.45%	11.11%
Jan 1981 - Apr 2017 Annualized Sharpe (Rf=4.22%)	0.44	0.55	0.35	0.50	0.49	0.51
Jan 1981 - Apr 2017 Worst Drawdown	-50.95%	-23.59%	-26.20%	-21.42%	-26.87%	-26.87%
20 Years
May 1997 - Apr 2017 Annualized Return	7.52%	10.18%	7.58%	8.15%	9.07%	9.17%
May 1997 - Apr 2017 Annualized Std Dev	15.17%	10.02%	10.98%	10.09%	10.38%	9.98%
May 1997 - Apr 2017 Annualized Sharpe (Rf=2.04%)	0.35	0.80	0.49	0.59	0.66	0.70
May 1997 - Apr 2017 Worst Drawdown	-50.95%	-15.28%	-17.37%	-16.18%	-15.28%	-15.28%
10 Years
May 2007 - Apr 2017 Annualized Return	7.00%	10.02%	6.93%	7.65%	7.50%	8.40%
May 2007 - Apr 2017 Annualized Std Dev	15.25%	9.18%	10.62%	9.86%	9.74%	9.41%
May 2007 - Apr 2017 Annualized Sharpe (Rf=0.52%)	0.42	1.03	0.60	0.72	0.71	0.83
May 2007 - Apr 2017 Worst Drawdown	-50.95%	-10.06%	-17.37%	-16.18%	-14.77%	-12.00%
5 Years
May 2012 - Apr 2017 Annualized Return	13.36%	11.91%	10.08%	10.84%	10.12%	10.58%
May 2012 - Apr 2017 Annualized Std Dev	10.19%	8.77%	9.55%	8.82%	9.29%	8.98%
May 2012 - Apr 2017 Annualized Sharpe (Rf=0.15%)	1.29	1.34	1.04	1.21	1.07	1.16
May 2012 - Apr 2017 Worst Drawdown	-8.36%	-7.64%	-14.78%	-9.88%	-14.77%	-12.00%

As we can see, relatively simple solutions work well to mitigate the impact of luck in rebalance timing.

Disclaimer

All content displayed here is for informational purposes only and is not guaranteed to be accurate, complete or up-to-date. Nothing herein should be considered investment advice or recommendation to buy, sell or hold any securities.

Building blocks of systematic investment strategies

Fri, 22 Jul 2016 00:00:00 +0000

It is important to understand the building blocks of systematic investing strategies before learning how to build them. Here is a schematic from the book, Inside the Black Box - The Simple Truth about Quantitative Trading by Rishi Narang, that provides a good way to visualize these building blocks and how they fit together in a system.

The Black Box Revealed

The author had aptly titled it - The Black Box Revealed, since that is how most people perceive it to be from the outside. From this schematic, it is easier to understand and visualize how a system is built with a systematic approach.

Here are the components of the system:

Data - Encapsulates the processes of raw data collection, cleaning and preparation in tidy form to be used by the models.
Alpha Model - Describes the core strategy, product universe (stocks, ETFs, futures) and the rules to be implemented.
Risk Model - Describes position sizes or how capital/risk is allocated. This is just as much important as the alpha model itself.
Transaction Cost Model - Describes the transaction costs by product. These may be theoretical numbers while backtesting the strategy using historical data and later adjusted based on live execution data.
Portfolio Construction Model - Describes how a portfolio is constructed after putting together positions by product across one or more rules of the core strategy.
Execution Model - Encapsulates a live trading system. Describes the tactical implementation of the core strategy.
Research - Sound research forms the underpinning of all models and is driven by quantitative and statistical analysis.

The interactions shown are typical, although there could be more depending upon the implementation. More than one model could also be combined together, depending upon how the strategy is formulated. Nonetheless, for most practitioners looking to build a complete system from scratch, this is a pretty good blueprint.