A 20% retail price increase would require almost a 40% steel price increase. With that being said, however, your 20% may be closer than most would think, but because of some other factors in addition to the steel increases - at least in the U.S. Also, that increase will take at least a couple of years to reach the retail customer.
There will be some offset to the steel increases in the US because of the repeal of the 30% tariff imposed in 2003. But only some offset because the price increases will be more than the tariff was.
One of the largest factors in US pricing is freight. The retail customer pays the freight indirectly on all phases of production. From the raw materials to final finished goods delivery to the farm. Ever notice how trucks outnumber cars on the interstate? The new regs on drivers and fuel increases yet to come will see freight start edging towards $3.00 per mile in the next few years (it averages around $1.25 now). It is already cheaper to bring 40,000 lbs of freight (a typical semi load) all the way from China to port in the US than it is to take that same container from port to the east coast! The cost of steel is only put into the product 1 time. The cost of freight is added any time a raw material, cut part, subassembly, or finished good is moved, whether that move be in processing, material handling, warehousing, distribution, or delivery to the customer. So while raw material inputs do contribute to retail price levels, it is the inputs that are added at multiple levels, like freight and labor, that really drive prices up quickly. I guess to conclude, I think you may be close on your estimate of a 20% increase at retail pricing, but in my opinion that increase will be from the combination of inputs that occur on multiple levels.