If variance of a residual (think: the error of regression models) is constant, then the variance is considered homoskedastic, which is fantastic if it’s your regression model.
Homoskedasticity implies that the data fits the regression model well, while its opposite...heteroskedasticity...implies that there’s some explanatory factor missing from the model (probably).
Homoskedasticity is an assumption that’s foundational to regression models. Many economists are regression model fans, since they have a way to try to measure cause-and-effect, along with the strength of each causing factor.
Related or Semi-related Video
Cost Accounting: What is Variance Analys...3 Views
And finance Allah shmoop What is variants Analysis Well basically
it's the fancy business Way to ask Why didn't things
go as expected Or how on Earth are we so
far off Plan plan is the key word You made
a plan a k a Budget Others in the company
rely in your planning and budgeting You know from the
hiring needs of the union factory workers this year to
the marketing spend for TV ads from that department to
the legal team who's fully plumbed up for the eleven
point three lawsuits you expected you'd have to defend this
year But then real life happened and things didn't quite
go as you expected What happened Well they'll figure it
out You're going to use variance analysis And yes it's
a whole science unto itself You own zesty kitty a
leading provider of condiments for pet food You launch a
new Suraj based sauce meant to pair perfectly with the
frozen mice that people feed to there Pet snakes It's
called spicy serpent surprise and cats love it As much
as I know a cat can love anything based on
market research and projections derived from other products you already
sell while you expect sales of three hundred thousand dollars
in the first month and you expect contributed profits from
this product to be one hundred twenty five grand giving
you a contribution margin of about forty two percent Well
a month after launch you look at the numbers you
brought in revenues of four hundred thousand dollars way better
than expected However profits and margins fell meaningful E short
of projections You only brought in one hundred thousand dollars
in contributed profits A contribution margin of just twenty five
percent Well shy of the budgeted forty two percent So
w t f baby Well you run some variance analysis
And as it turns out you had unexpected demand from
Ireland You were pretty sure you'd heard they didn't have
anything left there but oh well additional demand is usually
a good surprise Right That drove your four hundred thousand
over the much less revenue that you originally thought But
the extra demand meant you had to scramble to make
enough spicy serpents Surprised to feel all the orders Well
to do this you had to pay overtime to your
union workers in order to crank out the extra sauce
And that was expensive Now that you know what happened
you can well kind of adjust You know the extra
demand is there so you can hire some additional workers
carefully These new employees will get regular pay and the
additional capacity means you won't have to pay anybody usurious
overtime Since you're not paying the extra labor costs associated
with the overtime well margins will return to the expected
levels Meanwhile you adjust your revenue projections to the new
levels The added demand is still going to be there
next month so you revise your expectations You're going to
make them higher revenues go up So for the second
month you expect to repeat four hundred thousand dollars in
revenue However your labor adjustment should allow margins to get
closer to the originally expected forty two percent which would
give you contribution profits of about one hundred sixty eight
grand Right Well once the month is over you'll look
at the numbers and see if anything else doesn't match
the new budgeted expectations Then you'll run the variance analysis
process again and it becomes part of an ongoing cycle
Basically checking budgets against projections against riel Life results well
When the process is done you'll know whether your projections
were just wrong and need to be adjusted You know
like with the unexpected Irish demand for your new mouse
sauce or you'll find points in the production process and
distribution process and marketing process where your costs went awry
like they were too expensive to get Your stuff shipped
are marketed or placed or put on shelves or whatever
was needed and then you'll adjust You can implement changes
here to get things back to what you predicted Like
when you figured out how the overtime hurt your profit
margins and you brought in new workers overseas likely to
fix the unnecessarily high labor costs Well now thatyou're Suraj
a mouse sauce has taken off You can start your
RND department working on its next big project soy sauce
flavored insects for pets Spiders who
Up Next
What is covariance? Covariance is the comparison of how assets move in the markets. Positive covariance is when assets move in tandem, such as when...