4. Improving Budgets - Statistical Techniques Flashcards
What are the approaches for improving budgets?
1) Time series analysis
2) Linear regression
3) Index numbers
What is a times series analysis?
A series of figures recorded over a period of time. Most are made up of patterns.
What are the components of a times series?
Trend, seasonal variance, cyclical variance, random variation
What is a trend?
Underlying long-term movement in a constant direction, over a prolonged period of time
What is seasonal variance?
Predictable, recurring fluctuations over the short-term, typically over a year in duration
What is cyclical variation?
Recurring patterns like seasonal variations but tend to occur over a longer period of time
What is random variation?
Unpredictable fluctuations caused by random events
What is the additive model?
TS = T + SV
What is the multiplicative model?
TS = T X SV
How do you de-seasonalise data?
T = TS - SV
How can you use moving averages to identify trend figures?
- Find the trend - three day moving average
- Identify seasonal variations (TS-T=SV)
- Forecasting future figures - continue the trend and apply seasonal variation
How does linear regression help improve budgets?
It looks at what has happened in the past and assumes there is a reliable ‘linear’ relationship between factors.
What is the equation for linear regression?
Y=a+bX
Y= total costs X= volume of production outputs a= the point that the line intercepts the y axis which is the fixed cost b= the slope of the line which is the variable cost per unit
What are the assumptions and limitations of linear regression?
> Observations May not be typical of normal behaviour
Historical data may not be meaningful, patters change overtime
Assumes a linear relationship between variables
predicting outside of data range is less accurate
Reliability depends on how closely data line fits best fit
Total costs not affected by changes in volume
Affected by general rates of inflation
How can index numbers help improve budgets?
They measure the relative change in the volume or the value of an index over time
What is the ‘base year’?
The starting point for the index numbers and the year that we compare every other year to.
How do you work out the subsequent periods in index numbers?
(Current period figure / base period figure) X100
What is the retail price index?
It measures the changed its the prices of items of expenditure of the average household. It can be used to inflate or deflate costs at different points in time for more meaningful comparison so costs can be compared in ‘real terms’.
What do you calculate RPI?
E.g. (RPI2010 (base) / RPI2011) X ‘wage’ in 2011
What are flexible budgets?
When you prepare a variety of different budgets
What is regular re-forecasting?
Regularly revisiting and updating budgets
What are planning models?
Various software and spreadsheet models
What is the product life cycle?
The stages through which individual products develop overtime. It has four stages.
What are the stages for the product life cycle?
1) introduction stage: high costs and low revenue so still negative in cash flow
2) growth stage: falling unit costs and increasing revenue means that products should start to become profitable
3) maturity stage: high demand and sales but growth has slowed. Reduced production costs per unit with economies of scale and increasing efficiency. Product should be highly profitable.
4) decline stage: falling demand and market share leading to the product becoming loss making.
What is a census and what are its advantages and disadvantages?
When you look at every single item in the population.
Advantage - gain comprehensive information
Disadvantage - hugely time consuming, rarely used
What are the types of sampling?
Random sampling Stratified random sampling Systematic sampling Cluster sampling Multi-stage sampling Quota Census
What is random sampling and what are its advantages and disadvantages?
Samples are picked at random.
Advantages - quick and simple, random so should represent the whole population
Disadvantages - unrepresentative of wider pattern, entirely dependent upon the quality of sampling frame
What is stratified random sampling and what are its advantages and disadvantages?
The sample frame is divided into non-overlapping groups and a random sample is picked from each.
Advantages- represents all important groups, more representative of whole population, more meaningful information
Disadvantage - prior knowledge is needed, more time consuming
What is systematic sampling and what are its advantages and disadvantages?
Every nth item is taken.
Advantages - quick and easy, information is still random
Disadvantages - need access to whole population, regular patterns can cause issues, may not fully represent population.
What is cluster sampling and what are its advantages and disadvantages?
When the population is divided into groups/cluster and a number of groups is selected.
Advantages - quick and cheap, non-comprehensive frame required
Disadvantages- huge potential for bias, not fully representative
What is multi-stage sampling and what are its advantages and disadvantages?
Selecting a sample within each chooses cluster.
Advantages - Suits larger population, quicker to obtain sample
Disadvantages- not truly random, can be heavily biased
What is a quota and what are its advantages and disadvantages?
A sample of a certain number of items is taken.
Advantages - large population can be studied, simple and cheap
Disadvantages- may not represent entire population
How can historic data be unreliable?
It is old data and there is no guarantee that conditions will continue into the future.
How can unexpected events impact the reliability of forecasts?
They are impossible to factor in, assumptions are dangerous and reduce forecast reliability.
What do you need to consider when creating a forecast?
- How much data you have
- How reliable the data is
- Is there any data missing?
- is there any change in definition of the data?