Example 2 - Revision history

Rb: /* Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization) */

2014-11-11T13:54:44Z

‎Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization)

Rb: /* Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization) */

2014-10-30T14:24:06Z

‎Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization)

Rb: /* Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization) */

2014-10-30T12:03:06Z

‎Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization)

Rb: /* Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization) */

2014-10-30T12:02:12Z

‎Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization)

Rb: /* Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization) */

2014-10-30T12:01:51Z

‎Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization)

Rb: /* Recalculate statistics in subsamples */

2014-10-07T12:26:17Z

‎Recalculate statistics in subsamples

Admin at 14:29, 25 October 2012

2012-10-25T14:29:01Z

Admin: /* Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization) */

2012-10-25T14:09:45Z

‎Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization)

Admin: /* Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization) */

2012-10-18T13:57:28Z

‎Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization)

Admin: /* Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization) */

2012-10-18T13:46:35Z

‎Positive and negative returns (keywords: loops, if-then-else statements, logical operations, vectorization)

@@ Line 143: / Line 143: @@
 retminus = 0;
 for i = 1:T             %starts the loop
-     if lnrets(i)>=0
+     if ret(i)>=0
         Tplus = Tplus+1;            %counting non-negative returns
         retplus = retplus+ret(i);    %summation of non-negative returns

@@ Line 157: / Line 157: @@
 <ol>
 <li><p>Logical relationships also work for vectors, that is <source enclose="none">indpos = (ret>=0)</source> generates a vector of 0s (where <source enclose="none">ret(i)<0</source>) and 1s (where <source enclose="none">ret(i)>=0</source>)</p></li>
-<li><p>Logical expressions can be used for selecting subsamples from a sample, that is <source enclose="none">retplus=lnrets(indpos)</source> generates a subvector of non-negative returns and <source enclose="none">retminus=lnrets(1-indpos)</source> generates a subvector of negative returns.</p>
+<li><p>Logical expressions can be used for selecting subsamples from a sample, that is <source enclose="none">retplus=lnrets(indpos)</source> generates a subvector of non-negative returns and <source enclose="none">retminus=lnrets(logical(1-indpos))</source> generates a subvector of negative returns. Note that, while <source enclose="none">indpos</source> is a logical vector by definition, the result of <source enclose="none">1-indpos</source> is not any more, which is why we need the <source enclose="none">logical(1-indpos)</source></p>
 <source>indpos   = (ret >= 0);
 indneg   = logical(1-indpos); % the subtraction from one turns this into a double type

@@ Line 159: / Line 159: @@
 <li><p>Logical expressions can be used for selecting subsamples from a sample, that is <source enclose="none">retplus=lnrets(indpos)</source> generates a subvector of non-negative returns and <source enclose="none">retminus=lnrets(1-indpos)</source> generates a subvector of negative returns.</p>
 <source>indpos   = (ret >= 0);
-indneg   = logical(1-indpos); % the subtraction from one turns this into a double type, but we need a logical vector and hence the logical()
+indneg   = logical(1-indpos); % the subtraction from one turns this into a double type
 Tplus    = sum(indpos);
 Tminus   = sum(indneg);

@@ Line 159: / Line 159: @@
 <li><p>Logical expressions can be used for selecting subsamples from a sample, that is <source enclose="none">retplus=lnrets(indpos)</source> generates a subvector of non-negative returns and <source enclose="none">retminus=lnrets(1-indpos)</source> generates a subvector of negative returns.</p>
 <source>indpos   = (ret >= 0);
-indneg   = logical(1-indpos); # the subtraction from one turns this into a double type, but we need a logical vector and hence the logical()
+indneg   = logical(1-indpos); % the subtraction from one turns this into a double type, but we need a logical vector and hence the logical()
 Tplus    = sum(indpos);
 Tminus   = sum(indneg);

@@ Line 159: / Line 159: @@
 <li><p>Logical expressions can be used for selecting subsamples from a sample, that is <source enclose="none">retplus=lnrets(indpos)</source> generates a subvector of non-negative returns and <source enclose="none">retminus=lnrets(1-indpos)</source> generates a subvector of negative returns.</p>
 <source>indpos   = (ret >= 0);
-indneg   = 1-indpos;
+indneg   = logical(1-indpos); # the subtraction from one turns this into a double type, but we need a logical vector and hence the logical()
 Tplus    = sum(indpos);
 Tminus   = sum(indneg);

@@ Line 223: / Line 223: @@
 == Recalculate statistics in subsamples ==
-This part of the exercise is discussed on [[Example2b]].
+This part of the exercise is discussed on [[Example_2b]].
 =Footnotes=
   <references />

@@ Line 19: / Line 19: @@
 = Implementation =
 == Import data in MATLAB (keyword: data import) ==
@@ Line 73: / Line 75: @@
 GSKsimplerets = zeros(T,n);
 %This way r(1) = 0 by construction
-lnrets(2:end,:) = log(gsk_p(2:end,:))-log(gsk_p(1:end-1,:));
+GSKlogrets(2:end,:) = log(gsk_p(2:end,:))-log(gsk_p(1:end-1,:));
-simplerets(2:end,:) = gsk_p(2:end,:)./gsk_p(1:end-1,:)*100-1;</source>
+GSKsimplerets(2:end,:) = gsk_p(2:end,:)./gsk_p(1:end-1,:)*100-1;</source>
 You can also use a special command <source enclose="none">y = diff(x)</source> which generates a matrix <math>y</math>, such that <source enclose="none">y(i)=x(i+1)-x(i)</source>. This will result in a vector that is one row shorter than <source enclose="none">x</source> as from <math>T</math> observations we can calculate returns for periods <math>t=2,3,...,T</math>. In order to keep the return vectors in sync with the date vector we will therefore add a 0 return at the beginning, in place of the return at time <math>t=1</math>. This is the code we will use to calculate the returns. From now we will continue wil the log-returns only as this is common practice in applied Finance.
@@ Line 109: / Line 111: @@
 This is straightforward to implement by
-<source>numcorr = sum((gsk_r-GSKMean).*(app_r-GSKMean));
+<source>numcorr = sum((gsk_r-GSKMean).*(app_r-APPMean));
 dencorr = sqrt(sum((gsk_r-GSKMean).^2)*sum((app_r-APPMean).^2));
-corr  = numcorr/dencorr;</source>
+GSKAPPcorr  = numcorr/dencorr;</source>
 This could have been written in one line but was separated into denominator and numerator to make it a little easier to read.
@@ Line 128: / Line 130: @@
-A long way to compute these quantities is (where <source enclose="none">lnrets</source> is the relevant return vector, e.g. <source enclose="none">gsk_r</source>):
+A long way to compute these quantities is (where <source enclose="none">ret</source> is the relevant return vector, e.g. <source enclose="none">gsk_r</source>):
 <ol>
@@ Line 143: / Line 145: @@
      if lnrets(i)>=0
         Tplus = Tplus+1;            %counting non-negative returns
-        retplus = retplus+lnrets(i);    %summation of non-negative returns
+        retplus = retplus+ret(i);    %summation of non-negative returns
      else
         Tminus=Tminus+1;            %counting negative returns
-        retminus=retminus+lnrets(i);    %summation of negative returns
+        retminus=retminus+ret(i);    %summation of negative returns
 end
 retplus=retplus/Tplus;      %computing average non-negative return
@@ Line 155: / Line 156: @@
 <ol>
-<li><p>Logical relationships also work for vectors, that is <source enclose="none">indpos = (ret>=0)</source> generates a vector of 0s (where <source enclose="none">ret(i)<0</source>) and 1s (where <source enclose="none">ret(i)>=0</source>). <source enclose="none">indneg = (ret>=0)</source> achieves the opposite and is used to select the negative returns. You could have calculated <source enclose="none">indned = 1 - indpos</source>. However, this calculation turns the logical variable <source enclose="none">indpos</source> into a numeric <source enclose="none">indneg</source>. But for step 2 we need a logical variable which is why we repeat the logical operation just with a negative sign. An alternative way to get a logical is <source enclose="none">indneg = logical(1-indpos)</source>. Check out <source enclose="none">doc logical</source> to see why this works.</p></li>
+<li><p>Logical relationships also work for vectors, that is <source enclose="none">indpos = (ret>=0)</source> generates a vector of 0s (where <source enclose="none">ret(i)<0</source>) and 1s (where <source enclose="none">ret(i)>=0</source>)</p></li>
-<li><p>Logical expressions can be used for selecting subsamples from a sample, that is <source enclose="none">retplus=ret(indpos)</source> generates a subvector of non-negative returns and <source enclose="none">retminus=ret(logical(1-indpos))</source> generates a subvector of negative returns.</p>
+<li><p>Logical expressions can be used for selecting subsamples from a sample, that is <source enclose="none">retplus=lnrets(indpos)</source> generates a subvector of non-negative returns and <source enclose="none">retminus=lnrets(1-indpos)</source> generates a subvector of negative returns.</p>
 <source>indpos   = (ret >= 0);
-indneg   = (ret < 0);
+indneg   = 1-indpos;
 Tplus    = sum(indpos);
 Tminus   = sum(indneg);
@@ Line 176: / Line 177: @@
 <math>r_t = \phi_0 + \phi_1 r_{t-1} + \phi_2 r_{t-2} + \phi_3 r_{t-3} + \phi_4 r_{t-4} + \phi_5 r_{t-5} + \epsilon_t</math>
-the variables <source enclose="none">y</source> and <source enclose="none">x</source> are defined as (using <source enclose="none">gsk_r</source> as an example:
+the variables <source enclose="none">y</source> and <source enclose="none">x</source> are defined as (using <source enclose="none">gsk_r</source> as an example):
 <source>lags = 5;
@@ Line 205: / Line 206: @@
 fprintf(fid,'Mean(neg rets)      %6.4f     %6.4f \n',GSKretminus,APPretminus);
 fprintf(fid,'--------------------------------------- \n');
-fprintf(fid,'Corr(GSK_r,APP_r)        %6.4f \n',corr);
+fprintf(fid,'Corr(GSK_r,APP_r)        %6.4f \n',GSKAPPcorr);
 fclose(fid);</source>
 This will produce the following result:
@@ Line 219: / Line 220: @@
 Mean(neg rets)      -0.0136     -0.0220
 ---------------------------------------
-Corr(GSK_r,APP_r)        0.2143 </source>
+Corr(GSK_r,APP_r)        0.2143</source>
 == Recalculate statistics in subsamples ==
-This part of the exercise is discussed on [[Example 2b]].
+This part of the exercise is discussed on [[Example2b]].
 =Footnotes=
   <references />

← Older revision		Revision as of 13:46, 18 October 2012
Line 128:		Line 128:


−	A long way to compute these quantities is (where <source enclose="none">~~ret~~</source> is the relevant return vector, e.g. <source enclose="none">gsk_r</source>):	+	A long way to compute these quantities is (where <source enclose="none">lnrets</source> is the relevant return vector, e.g. <source enclose="none">gsk_r</source>):

	<ol>		<ol>