Forecasting Stock Prices in PHP

Question

Has anyone implemented stock prices forecasting, using php only.

Like we give data sets of 1 yr of open,high,low,close,volume and get prediction for next 15 or 30 days?

One example I saw is here https://stackoverflow.com/questions/13168568/how-do-i-use-holt-winters-seasonal-dampened-method-to-compute-a-two-month-sales

Just thought if we give a array of 1 to 10.. can next number be got - yes 11 is got, but couldnt go further than this? like getting next 5 data..

<?php
error_reporting(E_ALL);
ini_set('display_errors','On');

$anData = array(1,2,3,4,5,6,7,8,9,10);

print_r(forecastHoltWinters($anData));

function forecastHoltWinters($anData, $nForecast = 1, $nSeasonLength = 1, $nAlpha =     0.2, $nBeta = 0.01, $nGamma = 0.01, $nDevGamma = 0.1) {

// Calculate an initial trend level
$nTrend1 = 0;
for($i = 0; $i < $nSeasonLength; $i++) {
  $nTrend1 += $anData[$i];
}
$nTrend1 /= $nSeasonLength;

$nTrend2 = 0;
for($i = $nSeasonLength; $i < 2*$nSeasonLength; $i++) {
  $nTrend2 += $anData[$i];
}
$nTrend2 /= $nSeasonLength;

$nInitialTrend = ($nTrend2 - $nTrend1) / $nSeasonLength;

// Take the first value as the initial level
$nInitialLevel = $anData[0];

// Build index
$anIndex = array();
foreach($anData as $nKey => $nVal) {
  $anIndex[$nKey] = $nVal / ($nInitialLevel + ($nKey + 1) * $nInitialTrend);
}

// Build season buffer
$anSeason = array_fill(0, count($anData), 0);
for($i = 0; $i < $nSeasonLength; $i++) {
  $anSeason[$i] = ($anIndex[$i] + $anIndex[$i+$nSeasonLength]) / 2;
}

// Normalise season
$nSeasonFactor = $nSeasonLength / array_sum($anSeason);
foreach($anSeason as $nKey => $nVal) {
  $anSeason[$nKey] *= $nSeasonFactor;
}

$anHoltWinters = array();
$anDeviations = array();
$nAlphaLevel = $nInitialLevel;
$nBetaTrend = $nInitialTrend;
foreach($anData as $nKey => $nVal) {
  $nTempLevel = $nAlphaLevel;
  $nTempTrend = $nBetaTrend;

  $nAlphaLevel = $nAlpha * $nVal / $anSeason[$nKey] + (1.0 - $nAlpha) * ($nTempLevel + $nTempTrend);
  $nBetaTrend = $nBeta * ($nAlphaLevel - $nTempLevel) + ( 1.0 - $nBeta ) * $nTempTrend;

  $anSeason[$nKey + $nSeasonLength] = $nGamma * $nVal / $nAlphaLevel + (1.0 - $nGamma) * $anSeason[$nKey];

  $anHoltWinters[$nKey] = ($nAlphaLevel + $nBetaTrend * ($nKey + 1)) * $anSeason[$nKey];
  $anDeviations[$nKey] = $nDevGamma * abs($nVal - $anHoltWinters[$nKey]) + (1-$nDevGamma) 
              * (isset($anDeviations[$nKey - $nSeasonLength]) ? $anDeviations[$nKey - $nSeasonLength] : 0);
}

$anForecast = array();
$nLast = end($anData);
for($i = 1; $i <= $nForecast; $i++) {
   $nComputed = round($nAlphaLevel + $nBetaTrend * $anSeason[$nKey + $i]);
   if ($nComputed < 0) { // wildly off due to outliers
     $nComputed = $nLast;
   }
   $anForecast[] = $nComputed;
}

return $anForecast;
}
?>

random walk which is arima(0,1,0) with drift is the most accurate (according to several published articles in the literature) for forecasting stock prices. R's forecast has a function called rwf, you could check that out. — forecaster, Nov 29 '14 at 19:49
@forecaster Could you provide links for those articles? I'd love to read them— I'd always believed the naive forecast was the most accurate for stock data. — Zach, Nov 29 '14 at 19:56
@Ajmal You're question has been put on hold because it appears to be off-topic. [Take a look here](http://stats.stackexchange.com/help/on-topic), and then try re-wording your question. — Zach, Nov 29 '14 at 21:46
@Zach.. thanks for your concerns, this is the first q I have posted in this group, I wanted to learn if anyone has implemented forecasting in php? — Ajmal, Nov 29 '14 at 23:34
@Ajmal Probably not. PHP isn't a great language for statistical programming. You'd probably be better off doing the foresting in python or R (or another statistical programming language) and displaying the results in PHP. — Zach, Nov 30 '14 at 00:45
@forecaster Turns out what I've been referring too as the `naive` method is in fact a random walk forecast (without drift). The more you know! — Zach, Nov 30 '14 at 04:54

score 2 · Answer 1 · answered Nov 29 '14 at 19:45

2

It's very easy to forecast stock prices in any programming language. Given a 1-year dataset of OHLC data, use the most recent close as the forecast for the next 15/30/500 days.

This is called a naive forecast, and while it's not very exciting, it's more accurate than any other method I've ever seen for forecasting stock prices.

It's also trivially easy to implement.

answered Nov 29 '14 at 19:45

Zach

22,308
18
114
158

Do u have any sample code.. it would be helpful.. Thanks – Ajmal Nov 29 '14 at 19:46
@Ajmal I don't know PHP, but lets assume you have the 5 arrays of numbers: `open`, `high`, `low`, `close`, `volume` for each stock. If those arrays are sorted in date order, simply select the last element of the `close` array, and that is your forecast! – Zach Nov 29 '14 at 19:49

Forecasting Stock Prices in PHP

1 Answers1