[ 3 / biz / cgl / ck / diy / fa / ic / jp / lit / sci / vr / vt ] [ index / top / reports ] [ become a patron ] [ status ]
2023-11: Warosu is now out of extended maintenance.

/biz/ - Business & Finance

View post   
File: 16 KB, 954x202, corret.png [View same] [iqdb] [saucenao] [google]
1208940 No.1208940 [Reply] [Original]

I plan on getting into long-term investing and I've been playing around with some portfolio simulations.

However, I've been finding it more difficult than expected to find useful historical data. For example, as one of my tests I built a portfolio that includes:

SPY (S&P 500 index fund)
VBR (small cap value stocks fund)
STZ (sin stock, alcohol to be specific)
two REITs

I used the tool found at https://www.portfoliovisualizer.com/asset-correlations to get the correlations and other info on these assets, and it came up with some expected results and also some surprising ones (pic related).

- expected: low correlations between alcohol stock and other assets (low beta)
- expected: higher idiosyncratic risk on STZ
- expected: mid-to-low correlation between REITs and other assets
- surprising: the small cap value fund has smaller expected returns than the S&P fund; given that, as expected, it has a high correlation with the S&P one and a higher standard deviation, this makes itr a dominated asset. But I know that, historically, small cap value stocks have outperformed the market. So is the calculator wrong or is there an explanation I'm missing?
- surprising: the expected return on STZ is a whopping 49% annualized, which is ridiculous. Can someone explain this?

>> No.1208991
File: 55 KB, 474x480, 1419713979140.jpg [View same] [iqdb] [saucenao] [google]
1208991

Nobody? Too busy shilling coins?

>> No.1209026

Those aren't expected returns. They are historical returns. You can look at correlations of historical data, but doing optimizations based on it will give you pretty shit results.

>> No.1209031

>>1209026

Right, makes sense that they are historical returns. How is it possible for STZ to be as high as 49% though? Beta is only 0.83. Maybe the sample they take is too short to be accurate?

>You can look at correlations of historical data, but doing optimizations based on it will give you pretty shit results.

How would you go about it?

>> No.1210194
File: 1.85 MB, 300x222, 1423932467845.gif [View same] [iqdb] [saucenao] [google]
1210194

Bumpin for the dumpin

>> No.1210311
File: 11 KB, 909x295, STZsummarystats.png [View same] [iqdb] [saucenao] [google]
1210311

>>1209031

Here's summary statistics of STZ and ^GSPC via Yahoo using R and quantmod.

You notice that the value of 40 whatever percent is the average annual return over time. You also notice that the highest return was almost 100% and lowest was -6%.

>> No.1210312
File: 17 KB, 1280x1280, STZannualretplot.png [View same] [iqdb] [saucenao] [google]
1210312

>>1209031

If you plot the returns from the last couple of years you notice that one year was really high and it dominates the average. The S&P on the other hand has been much lower but steady.

>> No.1210314
File: 27 KB, 1280x1280, STZretplot.png [View same] [iqdb] [saucenao] [google]
1210314

>>1209031

This is what the plots of STZ and GSPC look like over time. It also shows their daily return relationship for the whole time period.

>> No.1210316
File: 8 KB, 960x480, STZbetaplot.png [View same] [iqdb] [saucenao] [google]
1210316

>>1209031

But remember that Beta is just the result of a regression and it depends on the data and it changes over time. Here is a plot of how STZ's 250 day beta has changed over the sample.

>> No.1210318
File: 37 KB, 1280x1280, STZtaplot.png [View same] [iqdb] [saucenao] [google]
1210318

>>1209031

Maybe you would prefer to look at a Technical Analysis plot and try and figure out STZ: My point is that it seems you don't have a grasp on what beta is, where returns come from, or what is the difference between historical and expectations (like someone else already mentioned). Be careful out there!