MSE Methods: Economic Metrics
MSE Methods: Economic Metrics Min-Yang Lee, NEFSC NMFS
Atlantic Herring MSE Workshop 2 December 7-8, 2016
Refresher
Economic objectives identified at the first workshop: Objective Maximize Yield for the herring Fleet Maximize Profit for the herring fleet Ensure herring catch temporal stability
Performance Metric Yield Net Revenues “Stability”
Performance Metric: Yield
Jon talked about this already.
Performance Metric: Net Revenues
• Build a simple model of the commercial herring industry • Net Revenue = Price*Yield-Cost
Price
• When landings go up,
prices go down. • When landings are very
low, consumers switch to menhaden. • Menhaden: $227 per mt at
the dock plus $133 per mt to transport. • Based on 2011-2015 data.
baseline
Yield
• Two fleets (Purse Seine and Trawl). • Divide the Yield historically (30% Purse). • Catch per day is similar for both fleets, but length of trips,
number of trips, and costs per day differ.
Costs
• Based on data on operating costs collected through
NEFOP. • Fuel, Damage, gear, ice, etc. • 2015 data only because...
Performance Metric: Net Revenues
Net Revenue = Price*Yield-Costs
Performance Metric: Stability
Many ways to think about the concept of stability: • How “streaky” is Net Revenue? • Do you get “runs” of bad outcomes and “runs” of good
outcomes? • How “persistent” are good and bad years? • If last year was a good year, is next year more likely to be a
good year? • “Booms and Busts” or “Steady”
Performance Metric: Stability
Many ways to think about the concept of stability: • How “streaky” is Net Revenue? • Do you get “runs” of bad outcomes and “runs” of good
outcomes? • How “persistent” are good and bad years? • If last year was a good year, is next year more likely to be a
good year? • “Booms and Busts” or “Steady”
Performance Metric: Stability
Many ways to think about the concept of stability: • How “streaky” is Net Revenue? • Do you get “runs” of bad outcomes and “runs” of good
outcomes? • How “persistent” are good and bad years? • If last year was a good year, is next year more likely to be a
good year? • “Booms and Busts” or “Steady”
Performance Metric: Stability
Many ways to think about the concept of stability: • How “streaky” is Net Revenue? • Do you get “runs” of bad outcomes and “runs” of good
outcomes? • How “persistent” are good and bad years? • If last year was a good year, is next year more likely to be a
good year? • “Booms and Busts” or “Steady”
Performance Metric: Stability
Many ways to think about the concept of stability: • How “streaky” is Net Revenue? • Do you get “runs” of bad outcomes and “runs” of good
outcomes? • How “persistent” are good and bad years? • If last year was a good year, is next year more likely to be a
good year? • “Booms and Busts” or “Steady”
An example of “Stable”
Another example of “Stable”
An example of “Streaky”
persistent good years
persistent bad years
Performance Metric: “Stability”
• For each simulation, we
can statistically test which of these we have. • For each operating model
(8x) and control rule (1,360), how many trials (out of 100) are “stable”? How many are “streaky”?
Streaky-ness and IAV are different
(a)
(c)
Low IAV, Stable
High IAV, Stable
(b)
(d)
Low IAV, Streaky
High IAV, Streaky
Streaky-ness and IAV are different
(e)
(g)
Low IAV, Stable
(f)
Low IAV, Streaky
High IAV, Stable
(h)
High IAV, Streaky
Atlantic Herring MSE Workshop 2 December 7-8, 2016
Refresher
Economic objectives identified at the first workshop: Objective Maximize Yield for the herring Fleet Maximize Profit for the herring fleet Ensure herring catch temporal stability
Performance Metric Yield Net Revenues “Stability”
Performance Metric: Yield
Jon talked about this already.
Performance Metric: Net Revenues
• Build a simple model of the commercial herring industry • Net Revenue = Price*Yield-Cost
Price
• When landings go up,
prices go down. • When landings are very
low, consumers switch to menhaden. • Menhaden: $227 per mt at
the dock plus $133 per mt to transport. • Based on 2011-2015 data.
baseline
Yield
• Two fleets (Purse Seine and Trawl). • Divide the Yield historically (30% Purse). • Catch per day is similar for both fleets, but length of trips,
number of trips, and costs per day differ.
Costs
• Based on data on operating costs collected through
NEFOP. • Fuel, Damage, gear, ice, etc. • 2015 data only because...
Performance Metric: Net Revenues
Net Revenue = Price*Yield-Costs
Performance Metric: Stability
Many ways to think about the concept of stability: • How “streaky” is Net Revenue? • Do you get “runs” of bad outcomes and “runs” of good
outcomes? • How “persistent” are good and bad years? • If last year was a good year, is next year more likely to be a
good year? • “Booms and Busts” or “Steady”
Performance Metric: Stability
Many ways to think about the concept of stability: • How “streaky” is Net Revenue? • Do you get “runs” of bad outcomes and “runs” of good
outcomes? • How “persistent” are good and bad years? • If last year was a good year, is next year more likely to be a
good year? • “Booms and Busts” or “Steady”
Performance Metric: Stability
Many ways to think about the concept of stability: • How “streaky” is Net Revenue? • Do you get “runs” of bad outcomes and “runs” of good
outcomes? • How “persistent” are good and bad years? • If last year was a good year, is next year more likely to be a
good year? • “Booms and Busts” or “Steady”
Performance Metric: Stability
Many ways to think about the concept of stability: • How “streaky” is Net Revenue? • Do you get “runs” of bad outcomes and “runs” of good
outcomes? • How “persistent” are good and bad years? • If last year was a good year, is next year more likely to be a
good year? • “Booms and Busts” or “Steady”
Performance Metric: Stability
Many ways to think about the concept of stability: • How “streaky” is Net Revenue? • Do you get “runs” of bad outcomes and “runs” of good
outcomes? • How “persistent” are good and bad years? • If last year was a good year, is next year more likely to be a
good year? • “Booms and Busts” or “Steady”
An example of “Stable”
Another example of “Stable”
An example of “Streaky”
persistent good years
persistent bad years
Performance Metric: “Stability”
• For each simulation, we
can statistically test which of these we have. • For each operating model
(8x) and control rule (1,360), how many trials (out of 100) are “stable”? How many are “streaky”?
Streaky-ness and IAV are different
(a)
(c)
Low IAV, Stable
High IAV, Stable
(b)
(d)
Low IAV, Streaky
High IAV, Streaky
Streaky-ness and IAV are different
(e)
(g)
Low IAV, Stable
(f)
Low IAV, Streaky
High IAV, Stable
(h)
High IAV, Streaky
