Moving Averages and Bollinger Bands

Question 1

Filtering Time Series Types

Answer 1

Low-pass filter - removes short-term fluctuations, but leaves the trend.
High-pass filter - removes the trend but keeps the fluctuations.

Question 2

Causal Filters

Answer 2

A filter is:
* causal if its output depends only on past and present inputs.
* non-causal if its output also depends on future inputs.

Question 3

Moving Averages (Rolling Average/Mean)

Answer 3

Type of digital low-pass filter.

Question 4

Types of Moving Averages

Answer 4

Three types of moving averages:
- simple, where weights are equal.
- weighted, where weights form an arithmetic progression.
- exponential, weights form a geometric progression.

Question 5

Simple v Exponential

Answer 5

• EMA responds more quickly to prices.
• EMA takes into account all past data; SMA takes into account only n most recent data points.
• EMA only needs the most recent EMA value to be kept; SMA requires all n most recent data points to be kept.

Question 6

Bollinger Bands

Answer 6

A standard technical analysis indicator built with a moving average and standard deviation. The bands define “high” and “low” price levels on a continual, relative basis.

Question 7

Mean Reversion with Bollinger Bands

Answer 7

A “mean-reversion” trading strategy uses Bollinger Bands to:
* buys when they suggest the market “oversold”;
* sells when they suggest the market “overbought”.

Question 8

Overbought/Oversold Strategy

Answer 8

• Long when price is below lower band line
• Short when price is above upper band line

Attempts to trade corrections when the market has “overshot”.

Question 9

Simple Moving Average Formula

Answer 9

sum(prices) / n
n is the number of periods.

Question 10

Weighted Moving Average Formula

Answer 10

sum(price x weight) / sum(weights)

Question 11

Exponential Moving Average Formula

Answer 11

EMA_t = smoothingFactor x price + (1 - smoothingFactor) x EMA_(t-1)