http://stockflix.tumblr.com/

Forecasts Archive

Wed, 23 Jan 2013 19:17:55 +0100

A short post to let you know about a less well-known area of the Stockflix site, which may indeed be useful for doing simulations and validating the forecast algorithms we use. We’re talking of the Samples area, a large collection of static pages where we have accumulated a forecast snapshots for many stocks, taken at a given date in the past.

If you are not currently logged into Stockflix you access the samples from the Forecast page by following the link:

Otherwise, if you are logged in, point the browser directly to stockflix.com/samples.html:

Here you can select any one of the stock markets we cover, which leads to the directory listing of available forecast snapshots, in alphabetical order based on stock tickers. Click on a ticker to see the frozen forecast:

The example above shows the forecast for Yum Brands as of May 14, 2012. Note that the graphics is not interactive, we store a static picture of the forecast chart on the date it was created. This is the forecast Stockflix would have shown on that date if you were logged in and asked for YUM ticker in the Forecast page.

At the time of writing we have archived more than three thousand forecasts over a period of eight months. From time to time - usually once or twice a month - we add new forecasts to the Samples area. Forecasts are extracted from Today’s Picks and therefore have strong reversal or continuation probabilities.

Stockflix Trading System on Collective2

Thu, 06 Dec 2012 11:56:18 +0100

In a previous post we have described some experiments we have done on backtesting Stockflix Today’s Picks within a limited universe of stocks.

Following the same approach a few weeks ago we launched a real trading system on Collective2.com, which is a lively community with over 20.000 systems and 64K members supporting automated trading on stocks, options, futures and forex.

The Stockflix trading system is identified as ‘Stockflix Nasdaq100 CP’ and you can find it here.

It should be clear even from its name that the trading system is limited to stocks which are part of the Nasdaq 100 index. Also, we only consider the Continue to Positive signals (as we did in our original backtesting experiments). The signals we provide are based on tickers that make it through to Today’s Picks, filtered by some additional heuristics on expected target price. According to our model, only one stock at a time is held.

Performance so far has been admittedly rather poor, but we have seen from our simulations that it may take several month for the strategy to stabilize. Keep an eye on the Nasdaq100 CP trading system and let us know what you think.

Expected Price

Fri, 02 Nov 2012 11:00:52 +0100

Based on feedback by our users, we have added a little tweak to improve the information displayed by stock forecast charts.

As you can see in the screenshot below, a stock forecast now includes the expected price the stock may reach based on the statistical analysis of pattern frequency as performed by StockFlix.

Avg. Expected is the average of different target prices weighted by respective frequency, while Return is the relative increase or decrease of this value with respect to the stock last price.

The same information is of course presented when you open a chart in the Today’s Picks page or in your Watchlist section of the Favorites page.

Note that you can easily calculate the expected price and return values based on the information provided by the interactive chart and the accompanying table, but it may be convenient to have these values at hand when browsing any stock forecast.

We have used the expected price - with some corrective factors - in the backtesting experiment we described in a previous post.

Backtesting Stock Picks

Tue, 02 Oct 2012 10:46:46 +0200

In this post we’ll describe some experiments that we have done recently on backtesting Stockflix recommendations. As you know, Stockflix has three major - highly correlated - functionalities to:

Identify the pattern exhibited by a given stock chart (and searching for stocks with similar patterns) [Search]
Forecast price movement scenarios of a specific stock by running a statistical analysis of the historical matches of the pattern that the stock currently has [Forecast]
Pick the most interesting stocks for the day, by sorting and ranking the results of calculating the forecasts for the whole stock universe and extracting those exhibiting a well defined behaviour [Today’s Picks]

Stock picking is certainly one of the most intriguing functions of Stockflix for potential investors. In this experiment, we wanted to simulate how ‘buying’ stocks picked by Stockflix algorithms would have performed over a given period of time, subject to certain conditions described below.

Model and Assumptions

As backtesting is a tricky subject which can get very complex, we have made a number of simplifying assumptions in our experiments.

Basically, we have chosen a limited stock universe, i.e. the components of the Dow Jones Industrial Average index (30 stocks), and a specific time period, from November 2007 to July 2012. Also, we only considered a single type of ‘positive’ signals as provided by Today’s Picks function, i.e. those grouped under the Continue to Positive tag in the service. The period considered includes both bullish and bearish stretches for the overall market.

Over that time period, using historical data we have extracted the signals that Stockflix would have provided for the Dow stocks. Given this is a small universe consisting of only 30 stocks, signals are not always available (on average, 20% of the dates in the considered time period show a signal).

Based on those signals, we simulated trading the stock in this pre-defined universe over a period of one year. As the signals we start from are positive we didn’t take into account short-selling strategies.

As we focused on a time period of over four and a half years overall, it turns out there are over 900 rolling years to consider. The following picture should give you the idea.

Therefore, for each year considered, starting from Nov. 15, 2007 to Nov. 15, 2008, and moving forward one day at a time, we ran a simulation of buying, holding and selling a stock list based on parameters provided by Stockflix.

We made a number of additional assumptions to simulate buying and selling stocks:

Buying and selling is done at end-of-day price
No trading costs are considered
Each list holds one stock at any given time

The model is summarized in the following example flow:

Buy and Sell Criteria

We adopted the following criteria for simulating buy and sell operations:

At any given day, buy a single stock; the selected stock is the one showing a Continue to Positive signal within the Dow Index constituents for that day; if there is more than one stock available, choose the one with higher Up probability
Sell the stock when either one of these conditions has been met: 1) the stock has reached a target price; the target price is calculated as a percentage (target factor) of the weighted average of Stockflix up probability forecast, or 2) the stock has reached a stop price; the price is calculated as a percentage (stop factor) of the weighted average of Stockflix down probability forecast
Keep the stock for at least w days before selling; w is calculated as a percentage of the average length of a leg of the pattern the stock had at the time it was selected

Here is an example of target factor calculation. Suppose we have this forecast:

The current price is 120.85. The continuation target price weighted by the upward (green) probabilities calculated by Stockflix is 145.7. The target factor is a percentage between 0 and 100 applied to the difference between the target and current price.

A target factor of 50% means that the stock is sold as it reaches 133.27 (i.e. an increase of 0.5 x (145.7 - 120.85) on current price). A similar definition holds for the stop factor, using the downward (red) probabilities.

We have introduced the target factor and the stop factor as we noticed that using a percentage of the target price as calculated by Stockflix instead of the plain average or a specific value yields better results overall.

Results

Given the above (admittedly strong) hypothesis, we have run a number of simulation batches, repeatedly varying the two parameters driving the target price and the stop price.

Each batch consisted of the above 900+ steps, each step in turn corresponding to trading the stocks in the Dow Jones universe for one specific year. The result of a simulation batch is how many times the Stockflix selection beat the corresponding Dow Jones Industrial Average index out of the 900+ years tested.

After extensive simulation, it turns out that the best parameter combination has a 70% target factor and a 30% stop factor. With this settings, lists based on Stockflix picks outperform the index in about 93% of the cases.

The following is a chart that compares the performance of Stockflix picks (in red) to the index (blue) for one random year with the above parameters:

Further testing should be done, particularly to extend the model to include multiple stock holdings, to analyze other signals such as Reverse to Positive, to test different stock indexes, to include trading costs, to carry out simulations over longer and shorter periods, etc.

However, these results provide us with increased confidence that the forecast method based on statistical analysis of pattern matching as modeled by the Stockflix service is indeed valuable and can provide to subscribers new investment ideas which are worth exploring.

New feature: Save to Favorites

Tue, 28 Aug 2012 14:35:00 +0200

By popular demand, we’ve added a little new feature to the Stockflix web site which allows you to save stocks to your personal Favorites page from anywhere in the site.

You’ll notice that when you open a forecast analysis chart a new Save to Favorites button appears at the bottom right of the chart.

When you click on the button, the stock is added to your Favorites list - without the need to go to the Favorites page first - and you’ll be notified with a pop up box that the operation has been completed:

The pop up disappears after a few seconds.

This functionality is available everywhere a forecast chart is opened, be it in the Today’s Stocks page or in the Forecast page. In this way you can follow your preferred flow of analysis and at the same time keep your stocks watchlist up to date without distractions. The watchlist can be accessed from the Favorites page as before, and you can still add stocks from there:

We hope you find this feature useful as we do.

A brief intro to Stockflix - Favorites

Thu, 26 Jul 2012 10:29:02 +0200

Stockflix’s Favorites page allows you to keep track of two sets of important items while analyzing stock trends and patterns over time:

Save a list of stocks for periodic forecast analysis (Watchlist)
Save a list of patterns and accompanying selections for future searching

After saving a few stocks and a few patterns your Favorites page might look like this:

Watchlist

If you have ever used the ticker input field in the Forecast page to request a forecast for a stock, this will look familiar. To add a stock to your watchlist, simply start typing its ticker or name in the text field and a pop up will appear with possible completions. Select the one you are interested in and the stock is added to the list.

Up to now you haven’t requested a forecast analysis yet. Once you click on a line in the watchlist, the forecast analysis for the selected stock is then computed on demand and displayed in a box below the line. These are exactly the same results that can be obtained by requesting the forecast from the Forecast page, but the watchlist saves you from moving between different pages, and allows you to check all your favorite stocks conveniently from one place.

Once you are no longer interested in tracking a stock you can simply delete the ticker from the watchlist by clicking the X icon on the right:

Patterns

This feature comes in handy when you have worked on a pattern by modifying its pivot points and maybe searching for its matching stocks within a specific sector or exchange. You can then save the pattern in its current form along with its related selections and have it ready in the Favorites page. This saves you a lot of manual work that should be done again to draw the pattern in the dashboard widget. On the other hand, if you only search for Classic Patterns then this feature might prove less useful.

Note that you save a pattern directly from the Search page, by clicking the “Save Pattern” button in the pattern drawing widget. When you save the pattern you give a name to it.

In the Favorites page, saved patterns are listed by their names. It is convenient to provide mnemonic names to patterns, which can help you remembering their purpose. Also note that a miniature but accurate image of the pattern shape is shown on the right of its name, along with the date and time of its creation. You can delete or rename a pattern by clicking on the respective icon to the right of the table.

When you click on a pattern name the associated saved pattern chart is loaded into the dashboard widget in the Search page, the search on the pattern is launched, and you are redirected to the results section within that page.

At this point you can further modify the pattern pivot points, filter on a specific exchange or industry sector, and maybe save the new configuration under a different name.

Remember that the Favorites page is reserved for Stockflix Premium users, if you have a free subscription you won’t be able to access these features. You are invited to sign up for a 7-day free trial to fully explore the available Premium functionalities.

A brief intro to Stockflix - Today's Picks

Thu, 12 Jul 2012 19:25:33 +0200

One of the most exciting features of Stockflix is its daily selection of hot stock picks. Here we’ll take a look at how these picks are organized and generated.

You can access the daily picks from the Today’s Picks page, after you’ve successfully signed in to Stockflix (a limited selection of picks is also available to guest users). Here is a snapshot of the page:

Daily Picks are grouped into four categories, respectively:

Reverse to Positive
These are the stocks most likely to invert their (most recent) negative price trend

Reverse to Negative
Stocks that according to Stockflix pattern matching algorithms have reached a peak and are more likely to move according to a downward slope

Continue to Positive
Stocks that will likely continue their positive run

Continue to Negative
These are the stocks with a negative forecast for the period

For each category, Stockflix shows up to ten stocks, ranked according to its algorithm. These are selected daily within the complete universe of over 20,000 stocks analyzed. Note that the actual forecast for a stock is shown ‘on demand’ as soon as you click on the line, it may take a couple of seconds to display.

You may also select a specific exchange - among those supported by Stockflix, including major US and European exchanges - to constrain the Picks to stocks listed on that exchange.

You have probably noticed that the forecast chart and table which are dynamically inserted in the Today’s Picks page are the same you could access from the Forecast page by entering the ticker or the company name there.

This fact is important to clarify the meaning of terms such as ‘most likely’ and ‘probable’ as used here, as of course exactly the same concepts apply: In Stockflix Forecast, price movement scenarios of a stock are calculated by running a statistical analysis of the outcomes of all the historical matches of the intrinsic pattern of the stock.

At an abstract level, Today’s Picks are selected by going through over 20,000 stocks in the Stockflix universe, calculating their respective Forecasts and identifying the intrinsic patterns with the most consistent set of outcomes based on the four categories above. This makes it possible to rank the stocks which are then presented in the page.

In theory, you could go through the universe of stocks one by one, collecting their forecasts and sorting and ranking the results manually to achieve a selection similar to that presented in the Today’s Picks page. Stockflix makes this process fully automated and convenient.

As an additional benefit, daily picks are also summarized and sent to our subscribers in a daily email digest.

Although we have collected significant data with respect to the four above categories that provide an excellent empirical support for the signals we summarize in the Today’s Picks selection, please remember that picks are not sell/buy recommendations!

The intent of Stockflix Picks is to provide you with a set of new investment ideas based on statistically sound analysis which can be used directly but are worth exploring with additional complementary tools such as fundamental analysis. A richer array of investment options means more diversified portfolios and ultimately reduced risk.

We have seen that stock forecasts can be requested explicitly for any stock of choice in the Forecast page, or implicitly by accessing the Today’s Picks selection. A third way to run a forecast analysis is by means of the Watchlist feature accessible through the Favorites page. This will be the subject of a future post.

A brief intro to Stockflix - Forecasts

Tue, 19 Jun 2012 10:13:10 +0200

In the previous post, we described pattern matching, one of the core functionalities of Stockflix. We have seen that when you expand a pattern matching result - that is, a pattern and the underlying stock price chart currently matching that pattern - you can click on the Forecast for this stock button from the chart in the Search page. This brings you to the Forecast page for that stock.

What is a stock forecast? The StockFlix view of the markets dynamics is based on pattern matching. So, for us at StockFlix, forecasting is a way to look into the future price movement scenarios of a specific stock by running a statistical analysis of the outcomes of all the historical matches of the pattern that the stock currently matches (this is called the ‘intrinsic pattern’ of a stock). Here is a snapshot of an intrinsic pattern:

Stockflix compares the intrinsic pattern to the patterns in its database to understand how it may evolve, which means it identifies possible values of its future pivot points. This information is presented in the interactive forecast chart and in the accompanying table. For instance, if you look at the following chart fragment:

you can see that the stock has a 10% probability to reach the price of 11.46 (an 8% increase on current price). This value is derived from observing the past behaviour of hundreds of similar patterns over several years of price data for thousands of stocks in different exchanges and keeping track of its frequency. This information helps you to improve your trading strategy by supporting the sizing and timing of your buy/sell/hold decisions.

The table to the right of the chart summarizes the frequencies of potential outcomes and gives an overall probability for the stock to reverse or continue its current trend as represented by its intrinsic pattern.

By hovering with the mouse on the blue dots in the forecast chart you can see how long it took for the stock price to move through adjacent pivot points:

In this case, it took 64 days to go from pivot at 12.84 price to next pivot at 10.01.

We have reached the Forecast page from the Search page by clicking on the ‘Forecast for this stock’ button, but of course you can request a forecast for any stock you like in the Stockflix universe simply by typing a company name or ticker in the input box of the page.

Search and Forecast are highly intertwined, and we’ll explore this relationship in the next post.

A brief intro to Stockflix - Patterns

Wed, 06 Jun 2012 16:42:20 +0200

In the next few posts, we are going do describe the major functionalities Stockflix offers to smart investors. These short descriptions, together with the FAQs and Approach pages, should help you get a better understanding of Stockflix.
Pattern matching is the core foundation of the service. You perform pattern matching from the Search page after you’ve signed in to Stockflix.

Even if our eyes seem to easily recognize patterns, pattern matching on price-time series has proven to be a difficult task. StockFlix is capable of ultra-fast and accurate pattern matching on price-time series. In the figure below, you can see a typical pattern on the left (Technical Analysis aficionados call it ‘Inverse Head and Shoulders’) and a matching stock chart on the right, where the pattern shape has been superimposed on the stock chart.

You can immediately recognize the matching, even if the pattern is tilted to the left. This example is an actual screenshot of the Dashboard widget available in the Search page.

A powerful feature of pattern matching as provided by Stockflix is that you can either select a pattern among those typically used in Technical Analysis (such as the one above), or you can draw your pattern in the Dashboard (and save it for later reuse). You can even use the price chart of a stock in the past and ask Stockflix to use that chart as a base pattern. We call this feature the ‘Time Machine’.

Below is a screenshot with some classic Technical Analysis patterns:

You can recognize the following patterns:

Cat and Dog
Inverse Cat and Dog
Upward and Downward Channel
Head and Shoulders
Inverse Head and Shoulders
Downward or Upward Break
Simple swing (down or up first)

Here is a snapshot of the Time Machine.

Usage is simple, just type the name of a company or a ticker and choose a date in the past (e.g., four months ago). Stockflix will immediately search the selected stock price chart and recognize the intrinsic pattern that the stock chart had four months ago. This pattern is shown in the dashboard widget and you can further modify it according to your preferences.

The Search page is organized around three tabs, corresponding to the above workflows. You can also mix and match the steps according to your taste, e.g. you can start from a typical T.A. pattern such as Head and Shoulders, and interactively modify it to your preference.

Independently of your approach, as soon as a pattern has been drawn in the dashboard widget, Stockflix is able to search across thousands of price charts in a few seconds and show the results. Each result is displayed as an interactive graphics showing the latest stock price chart with the pattern superimposed on it.

Results are ordered according to the quality of their match (best matches first). The quality of matching is represented by stars in the result line. As soon as you click on a symbol the actual price chart with its corresponding pattern match is opened inline. By hovering with the mouse on a pivot point the date of the pivot and the closing stock price is shown on the dashboard.

Clicking on the ‘Forecast for this stock’ brings you to the Forecast page, the subject of another article. Note that you can restrict the search to a specific exchange or industry sector by filtering the results with the option menus in the right panel.

The Search page favors an interactive, exploratory style to pattern matching. You usually start from a classic pattern or from the Time Machine and progressively tweak the pattern by hand - adding or modifying pivot points - until you find interesting results. At this point you can save the pattern, possibly with accompanying exchange and industry sector selections. You find all your saved patterns in the Favorites page (please note that saving patterns is a feature available to Trial and Premium users only).

Now that you are able to screen a huge number of stocks with ease, speed, and flexibility based on visual matching, it is time to look at the forecast analysis capabilities of Stockflix. This will be the topic of one of our next posts, check back soon.

Yes We're Open

Fri, 30 Mar 2012 18:55:00 +0200

In the past several months, we have worked hard on an innovative financial site that we are launching today: stockflix.com.

The basic idea of StockFlix is to provide what we call stock chart intelligence and put it right into the hands of smart private investors. Check it out and let us know if you like it!

In the next few posts we’ll provide some insights into Stockflix innovative pattern matching technology and explain how to best take advantage of it.

Stay tuned.