周五高盛的讲座，关于Behavioral Finance

恩，比上周的那次有趣很多

换乱的记了一点点东西，贴在这里备查

----------------------------------------------------------------------------

Behavioral Finance:

1. Introduction and Background

2. A Closer Look at Behavioral Finance

3. Applications in Asset Management

4. Application in Optimal Investment Policy

Neoclassical Finance

Assumptions:

1. Rational Agents

2. Good and perfect information

3. Frictionless market

Predictions

1. Efficient market

2. Diversified risky portfolios

3. Irrelevance of institutional forms.

Behavioral Finance helps to better understand:

1. Investor behavior

2. asset pricing

3. Managerial behavior

Common Biases:

1. Representativeness/ overreaction

一个比较有趣的例子，分解成Bayes rule,说人可能会对其中的P(A|B)项overreact。

Confident should depend on the size of sample/

But overreaction make people neglect size of sample.

2. Conservatism/underreaction

Box 1: 30% red, 70% blue;

Box 2: 70% red balls, 30% blue

One box is chosen randomly, and out of the 12 balls chosen from it, 8 are red and 4 are blue, what is the probability that balls where drawn from Box 2?

Ans: 98%!!!

Key: 完全没有信息的时候，得到一个信息，容易过度反应。

有大量的先验知识的时候，容易underreact----市场上得到比较好的消息的时候，有时候会反映迟钝

3. Available Heuristic

4. Overconfidence

在瑞士90%的人认为自己的above average-----在股市上表现为很多人交易量很高

5. Narrow Framing

We look at a portfolio in isolation

题目Ｉ: which do you prefer:

ａ: 稳赚７６００

ｂ: 25%的概率不赚，７５％的概率赚１００００

听说一般人都选b----在只有盈利，没有损失的时候，一般人比较愿意ｔａｋｅ　ｓｏｍｅ　ｒｉｓｋ

题目II：

　　c: 赔２５００

　　d: 25％的概率赔10000,75%的概率不陪

一般人选c------在有损失的时候，很多人不愿意再有不确定因素

但是如果放在一起考虑，

　　　ａ＋ｄ：２５％赔２４００，７５％赚７６００

　　　ｂ＋ｃ：２５％赔２５００，７５％赚７５００

怎么看都是ａ＋ｄ比较好－－在孤立的看问题的时候，容易产生错觉

这是一个非常有趣的例子，虽然不太make sense－－－at least to me.

6. Loss Aversion/Disposition Effect

Prospect theory:

We care about gains and losses( vs. just final wealth)

We are lose averse

We overweight small probabilities ( e.g. lottery tickets)

Limited Arbitrage:

Correlated investor demand + limited arbitrage = price effects

Noise trader risk (interim price risk)-----那帮疯狂的sb小股民

Fundamental risk (or arbitrage model risk) -------类似当年抄马克的那帮可怜人

Implementation costs

某年Palm股票的事情，因为没有办法short Palm的股票，所以arbitrage看着很大，没法运作….

Delegated arbitrage and short horizon:

-----LTCM当年就是这么挂的，借钱太多，短期顶不住，挂了~~~-----索罗斯用自己的钱就没有这个问题，唉，有钱就是nb呀

用来做Asset Manage：

Portfolio Construction Methodology

1. 用BF作一些信号，和传统的手段一起使用，构建一些长期portfolio

Some technique

1. 引入短期(6-12month) momentum-----一般人的短期记忆在这个时间段

2. Long-term reveral (3-5 year , good performers start to do poorly)

3. Value premium (low PB stocks outperform)

4. Size( small stocks outperform)

….

根据以上标准，选一半最好的股票，outperform index。可以买好的一般，short index。

用来做控制员工：（风控）

1. 发现员工长期选股过于集中

2. trading 太多

3. disposition effect (sell decisions)------买涨的股票一方面不符合momentum，另外盈利会导致税收。

4. buying decisions( representativeness)

Avoid making mistakes：

1. Avoid overconfident

2. Accept the limits of your knowledge

3. Do not extrapolate recent behavior while underweighting long-term averages.

4. Accept your losses gracefully.

Tricks to avoid these biases:

Rephrasing the question (use the framing effect)

Put the question in a visual instead of verbal format( avoid abstractions)

Standard evaluation procedure are helpful

Do your sells outperform your buys

Do you have lots of sells right around the break-even point (if so, disposition effect)

Are you earning back your training costs ( overconfidence) .

Taking advantage of other’s mistakes

1. …

2. if you can afford the short-term…

3. try to find out what retail investors are doing, then do the opposite

4. Beware of investment sentiment, especially if you are a contrarian