DAX+0,10 % EUR/USD+0,26 % Gold+0,13 % Öl (Brent)+1,11 %

Meine kleine Sammlung an Börsenstatistiken (Seite 9)


WKN: A0AET0 | Symbol: DE000DB2KFC6
2.748,50
21.06.18
Deutsche Bank
-0,71 %
-19,75 PKT

Begriffe und/oder Benutzer

 

Antwort auf Beitrag Nr.: 58.002.426 von faultcode am 17.06.18 01:39:08
Calling a company ‘great’ doesn’t make it a good stock (c) -- Regression to the Mean (1)
Obiges (*) Papier gibt breiten Raum dem Konzept von: Regression to the Mean

So auch hier besprochen: http://garysmithn.com/blog/peyton-manning-is-likely-to-regre…

=>
There are few statistical facts more interesting than regression to the mean for two reasons.
First, people encounter it almost every day of their lives. Second, almost nobody understands it

The coupling of these two reasons makes regression to the mean one of the most fundamental sources of error in human judgment, producing fallacious reasoning in medicine, education, government, and, yes, even sports.—anonymous journal referee



=> zurück zu (*):
hier der (mMn) entscheidende Satz zur richtigen Deutung und Anwendung:

What regression does imply, however, is that imperfect measurement of an unobserved trait tends to overstate the extent to which the underlying trait varies from the mean.

=> also (in meiner Übersetzung):

Was Regression (~) (in der Statistik) (dennoch) besagt, ist, dass eine mangelhafte Messung einer nicht beobachteten Eigenschaft dazu tendiert, das Ausmass der Variabilität dieser Eigenschaft um ihren Mittelwert zu überschätzen.

Beispiel:
• ein Unternehmen mit einem Gewinnwachstum, welches hoch ist im Vergleich zu einer Gruppe von Unternehmen, ist auch wahrscheinlich hoch gegenüber der Fähigkeit des Unternehmens für (hohes) Gewinnwachstum. (FC: verstanden? :D )

=> Gary Smith geht dann noch dabei interessanterweise auf eben diese Beobachtung bei Fama und French ("earnings regression") ein, deutet es aber im GGs. zu diesen rein statistisch:

=> d.h. Gary Smith deutet eine Regression to the Mean (z.B. beim Gewinnwachstum eines Unternehmens) rein statistisch!

--> er führt dazu Studien an (da mache ich morgen weiter... schon spät nun...)


Kahneman (und Tversky) gehen auf diesen Effekt auch näher ein --> Daniel Kahneman (2011): Thinking, fast and slow --> Kapitel 17 Regression to the mean und Kapitel 18 Taming Intuitive Predictions --> https://www.wallstreet-online.de/diskussion/1258587-41-50/me…



(~) https://de.wikipedia.org/wiki/Regression_zur_Mitte
Antwort auf Beitrag Nr.: 58.002.456 von faultcode am 17.06.18 02:55:25
Regression to the Mean vs. the Long view
Einschub.

Stock market risk is always far higher than you think
https://www.marketwatch.com/story/stock-market-risk-is-alway…


...The longstanding assumption that risk declines as we focus on longer and longer time horizons is, in essence, the notion that there is regression to the mean — that good periods are more likely than not to be followed by poor ones, and vice versa. And there’s no doubt that this is a powerful force. Over the last two centuries, the range of stock market outcomes at the one-year horizon was a lot greater (46 percentage points) than it is at the 30-year horizon (9 annualized percentage points).

To be clear, Professors Pastor and Stambaugh don’t deny that regression to the mean exists in the financial markets, or that it is a powerful force. But their argument is that there are other forces that are even more powerful and that, therefore, more than offset it.

How much does stock market risk grow as time horizon expands, once you take both regression to the mean and these other forces into account? The accompanying chart provides the answer: When moving from a one-year horizon to a 10-year horizon, risk (as measured by the variance of possible returns) expands by around 10%. When moving from a one-year to a 30-year horizon, this variance ratio expands by nearly 50%:



(FC: risk ist natürlich zweiseitig: negativ und postiv! Auch oben ;) )

If you have a hard time believing that, just remember the Japanese stock market over the last three decades. The Nikkei 225 Index NIK, +0.61% is currently trading for barely 50% of where it stood 30 years ago. In essence, the professors are arguing, don’t be too confident that the U.S. stock market couldn’t suffer the same fate.

For most of you, however, I bet that in your heart of hearts you already know that the professors are right. We’re unlikely to lose our job tomorrow, just as it is unlikely that the stock market will fall by 50%. But our confidence drops significantly when we expand our focus to what might happen over the next decade — or three decades, for that matter.

Risk grows the further into the future we focus, rather than becoming less.

The investment implication for retirees? You probably should reduce your equity allocation. This is not a market timing judgment, since this implication would be the same regardless of where the stock market stands. But you might find this advice particularly compelling in light of the advanced age of the current economic recovery.


Beitrag zu dieser Diskussion schreiben