In the last article marginal cost (“Grenzkosten”) told us what one more piece costs. Here it is used to make decisions: how much to produce, down to which price to keep working, how much to order at a time.
A workshop makes wooden pallets and sells them at 30 €. The market sets this price: there are many sellers, and it does not depend on any one workshop. How many pallets a month should it make? The owner compares three plans. For each he works out what one more pallet would cost and how much profit is left:
Pallets a month
300
400
500
One more pallet costs about
26 €
30 €
34 €
Profit a month
2200 €
2400 €
2200 €
At 300 pallets the next one costs 26 € and sells for 30: it pays to make it. At 500 the next one already costs 34 €, more than the price, and profit is falling. The best point is where one more pallet costs exactly what it sells for: at 400, with a profit of 2400 €. The rule “price equals marginal cost” solves the first task of the article.
In every task of this article a cost function is given, and the question is one of three:
how much to produce at the market price, as with the pallets;
down to which price to produce at all: in the short run (the shutdown point, “Betriebsminimum”) and in the long run (the optimum point, “Betriebsoptimum”);
how much to order at a time so that ordering and storage together cost least (the optimal order quantity, “optimale Bestellmenge”).
Three tasks, one cost function. The answer lies in a different place each time: where marginal cost equals the price, where average variable cost is lowest, and where ordering and storage together cost least. The overview puts the three paths side by side:
The market sets the price p (“Marktpreis”), and a cost function is given. Producing more pays while one more unit costs less than the price, up to where p equals K′(x).
p==K′(x)
Trap:taking the smaller of two solutions, where profit has a minimum; the right one is the larger, where K′(x) already rises.
The price floor (“Betriebsminimum”) or the lowest cost per unit (“Betriebsoptimum”). The fixed cost K(0) is paid even at a standstill, so the floor lies at the minimum of the variable cost per unit kv(x).
kv(x)==xK(x)−K(0)→→min
Trap:taking the minimum of the marginal cost; you need the minimum of kv(x), and the answer is a price, not a quantity.
A good is sold evenly all year and ordered in batches. The best order size x∗ depends only on the yearly sales λ, the cost of one order K and the yearly storage of one unit h.
x∗==h2Kλ
Trap:putting the purchase price per unit into the formula; x∗ contains only λ, K and h.
How much to produce at the market price: maximum profit
The workshop looked for the output at which one more pallet costs exactly 30 €, and tried out plans to find it. The formula finds the same place without trying. The exam task below asks for the profit at that output:
Exam-style task
Eine Genossenschaft verkauft Kürbisse zum Tagespreis von 55 GE pro Kiste, egal wie viel sie anbietet. Die Kosten für x Kisten betragen
K(x)==0.5x2++34x++70
Welchen Gewinn kann sie höchstens erzielen?
a)226.4
b)1155.0
c)150.5
d)220.5
e)178.6
How to recognise the task. The firm sells at a price it cannot influence: “Marktpreis” (market price), “Weltmarktpreis” (world market price), “vollständige Konkurrenz” (perfect competition), “viele Anbieter” (many suppliers). The cost function K(x) is given, quadratic or of third degree. You are asked for the quantity with the largest profit (“Gewinn”) or the profit itself.
The route to the answer in short; why the price is set equal to the marginal cost is explained in the “Why?” block under the formula:
Steps: Price equals marginal cost → Solve the equation → Choose the right solution → Work out the profit.
Formula sheet (“Formelsammlung”, the printed sheet): revenue and profit as functions of the price; a test for the maximum is not on it.
Memorise:p=K′(x), the solution with K′′(x)>0.
p=K′(x)
Let us solve the pumpkin task. “However much it offers” means the price of 55 GE (Geldeinheiten, monetary units that work like euros) per crate is fixed. The question here is not the quantity but the profit at it, so one more step is needed at the end.
K′(x)=x+34, the price is 55 per crate: x+34=55.
x=55−34=21 crates.
K′′(x)=1>0: marginal cost rises, and beyond 21 crates every further crate costs more than the price.
Revenue 55⋅21=1155, cost K(21)=0.5⋅441+34⋅21+70=220.5+714+70=1004.5. Profit 1155−1004.5=150.5 GE, option (c). Two neighbouring options appear when a piece is missing: 1155, option (b), is the revenue alone, and 220.5, option (d), leaves out the fixed cost of 70. The fixed cost does not move the quantity, but it belongs in the profit.
Another firm: it has the cost K(x)=0.02x2+14x+800 €, and the market price is 30 €. At what quantity is the profit largest, and how large is it?
K′(x)=0.04x+14, the condition 0.04x+14=30.
0.04x=16, x=400.
There is one solution, K′′(x)=0.04>0: this is a maximum.
K(400)=3200+5600+800=9600, G=30⋅400−9600=2400 €.
Cost of third degree. Then p=K′(x) is a quadratic equation with two solutions:
one solution is negative: with K(x)=0.25x3−3x2+19x+60 and the price 34, multiplying 0.75x2−6x+19=34 by 34 gives x2−8x−20=0, x=10 or x=−2; the negative one drops out, K′′(10)=1.5⋅10−6=9>0, the profit G(10)=340−200=140 €;
both solutions are positive: at the price 10 they are 2 and 6; at 2 the marginal cost is still falling, K′′(2)<0, and the profit has a minimum there, while the maximum is at 6 (there G(6)=−60, exactly like G(0): the price 10 is the shutdown point).
If the best profit is no higher than G(0)=−K(0), producing pays no better than standing still; that is what the section “Down to which price to produce” is about.
Trap. Taking the smaller of the two positive solutions: at the price 10 that is 2 instead of 6. Another mistake: forgetting the fixed cost in the profit.
Your turn1 of 3
A firm takes the market price as given (“Preisnehmer”, price taker); its cost is K(x)=0.1x3−1.5x2+10x+80, the price 7.6 €. At what quantity is the profit largest? Round to two decimal places.
Down to which price to produce: shutdown point and optimum point
As long as the price was given, the quantity was what you looked for. When the market price falls, the firm faces a different question: below which price producing is worse than standing still.
How to recognise the task. “Betriebsminimum” (shutdown point), “kurzfristige Preisuntergrenze” (short-run lowest price): down to what price the firm still produces. Or “Betriebsoptimum” (optimum point), “langfristige Preisuntergrenze” (long-run lowest price), “minimale Stückkosten” (smallest unit cost). You are given K(x), or the variable cost Kv(x) and the fixed cost separately.
In the short run the fixed cost K(0) is due even at a standstill, so the firm produces while the price covers the variable unit cost (“variable Stückkosten”):
kv(x)=xK(x)−K(0)
The shutdown point is the smallest value of kv, so it is a price. Some tasks also ask for the quantity x where it is reached, so read the question carefully.
Steps: Drop the fixed cost → Form the variable unit cost (without the fixed cost, per piece) → Find the minimum → Check with the marginal cost.
Formula sheet:pmin=V(xmin) (lowest price): V is the variable unit cost, our kv; how to find xmin is not on it.
Memorise: the shutdown point is the smallest value of kv(x)=xK(x)−K(0); there K′(x)=kv(x).
The cost is the same as in the article on marginal cost: K(x)=0.25x3−3x2+19x+60. Where is the short-run lowest price? The picture below shows the variable unit cost with the minimum 10 at 6 pieces, dashed the unit cost with the fixed cost, and the marginal cost curve, which passes through the minimum. The dot 18.50 on the dashed curve marks the minimum of the unit cost with the fixed cost. It does not answer the question about the lowest price. It is needed further down, at “The optimum point”.
K(0)=60, the variable cost Kv(x)=0.25x3−3x2+19x.
kv(x)=0.25x2−3x+19.
kv′(x)=0.5x−3=0 at x=6, kv(6)=10 €.
K′(6)=27−36+19=10, it matches.
The optimum point. The smallest unit cost with the fixed cost, k(x)=xK(x), is the long-run lowest price (“Betriebsoptimum”). You find it the same way, from k′(x)=0. An example: K(x)=x2+28x+9, k(x)=x+28+x9, k′(x)=1−x29=0 at x=3, and k(3)=3+28+3=34 €. For the cost K(x)=0.25x3−3x2+19x+60, the optimum point is at x≈7.92, k≈18.50: that is the minimum of the dashed curve in the shutdown-point picture above.
Trap. Taking the minimum of the marginal cost: K′(4)=7 lies earlier and lower. Giving the quantity 6 instead of the price 10. Taking k(x) with the fixed cost when the short-run limit is asked for.
Your turn2 of 3
A firm has the cost K(x)=0.1x3−3x2+40x+500. What is the shutdown point, that is the short-run lowest price? Round to two decimal places.
How much per order: cost-minimising order quantity
Both earlier tasks were about output. The last one is about buying in: the trader makes nothing, but pays for every order and for every day the goods sit in the warehouse. The task looks like this:
Exam-style task
Ein Bürofachhandel verkauft pro Jahr gleichmäßig λ=5550 Kartons Kopierpapier zu je 115 GE. Jede Bestellung beim Großhändler kostet K=555 GE, die Lagerung eines Kartons kostet h=5 GE pro Jahr. Wie viele Kartons sollte der Händler pro Bestellung ordern, damit die Kosten minimal sind?
a)612
b)785
c)1570
d)1110
e)750
How to recognise the task. A trader sells λ pieces a year at an even rate, each order costs K (here K without an argument: the cost of one order), storage h per piece and year, purchase c per piece. You are asked for the size of one order (“Bestellmenge”, order quantity), the number of orders a year or the days between two orders. The cost function is sometimes given as a hint (“Hinweis”):
C(x)==xKλ++cλ++2xh
Four steps; below them, the formula and where it comes from:
Steps: Match the quantities → Set up the formula → Put in and work out → Round and check.
Formula sheet:C(x)=xKλ+cλ+2x⋅h, with the cost K of one order; its minimum is not on it.
Memorise:x∗=2Kλ/h, without c.
x∗=h2Kλ
Let us solve the office supplies task. The purchase price 115 does not enter the formula: it is the same for every order size.
λ=5550 boxes a year, K=555 per order, h=5 per box and year.
x∗=52⋅555⋅5550=1232100=1110.
At 1110 boxes the ordering costs 1110555⋅5550=2775 equal the storage costs 21110⋅5=2775: that is how the optimum looks. The answer is 1110 boxes.
A trader has λ=20000, K=100 €, h=4 €, c=5 €. At what order quantity is the cost smallest? In the picture the dashed curve “Ordering” xKλ=x2000000 falls, the line “Storage” 2xh=2x rises, and the curve “Total” has its minimum at 1000 pieces.
λ=20000, K=100, h=4; c=5 is not needed.
x∗=h2Kλ.
x∗=42⋅100⋅20000=1000000=1000.
Ordering cost 10002000000=2000 €, storage cost 21000⋅4=2000 €: the same.
The tasks also ask these questions:
storage cost in percent: “8 % des Einkaufspreises pro Jahr” (8 % of the purchase price per year) means h=0.08⋅c; this is the only thing c is needed for;
number of ordersn=λ/x∗, for the trader in the second example 20000/1000=20; days between two orders360/n or 365/n, as the task says, for that trader 360/20=18;
a parameter asked for: if x∗ is known, the formula gives h=x∗22Kλ or K=2λhx∗2; for example x∗=800 gives the value h=6.25, and h=4 with x∗=1000 gives K=100.
Trap. Adding c to h, or forgetting the two under the root. Check: at the optimum the ordering cost and the storage cost are equal.
Your turn3 of 3
A trader sells λ=36000 pieces a year. Each order costs K=50 €, storage h=2.5 € per piece and year, purchase c=4 € per piece. What is the cost-minimising order quantity?
Weighted mean: the sum wᵢ(xᵢ − m)² at its smallest
In the order quantity two kinds of cost pulled the answer in opposite directions, and the best one lay in between. The question of where to put a warehouse is solved the same way: where it should stand so that the sum of weighted squared distances to the branches is smallest. It needs the same derivative, set to zero.
Formula sheet and what to memorise
All three tasks of the article have formulas on the sheet, only with different letters from those in the tasks. What to do with those formulas you have to remember yourself.
On the formula sheet:
the sheet writes the cost C, the revenue R(p)=p⋅D(p) (price p times demand D(p)) and the profit π(p); the task texts write K, E and G
revenue and profit as functions of the price: section “How much to produce at the market price”
the lowest price pmin=V(xmin), with V as our kv: section “Down to which price to produce”
the cost function of ordering and storage C(x)=xKλ+cλ+2x⋅h, with the cost K of one order: the section “How much per order” differentiates it
Memorise yourself:
market price: p=K′(x), the solution with K′′>0; work out the profit with the fixed cost (section “How much to produce at the market price”)
shutdown point: the smallest value of kv(x)=xK(x)−K(0), the answer is a price; optimum point: the smallest value of xK(x) (section “Down to which price to produce”)
order quantity x∗=2Kλ/h; at the optimum ordering and storage cost the same; the interval is 360/n or 365/n days (section “How much per order”)
location with the smallest sum ∑wi(xi−m)2: m=∑wi∑wixi