Investing in a single 10K Vishay metal foil resistor is worth it and it is just about in your budget.
If you have one very accurate resistor, it is possible to make your own precisely calibrated resistors that are multiples of 2 or ten different from the 10K (ie 100, 1K, 100K ,1M, 20K, 5K, etc) by building a Hamon divider and by making yourself a good microvolt galvanometer with a low current low offset opamp.
This is an exception resistor for $27:
http://www.digikey.com/product-detail/en/Y145310K0000V9L/Y1453-10KA-ND/2609894+/- 0.2ppm/C +/-0.005% accuracy . It should have fabulous long term stability.
With the Hamon divider, with care you can multiply or divide the values to at least 0.001% accuracy comfortably (relative to the 10K Vishay). Digikey also have some nice Vishay low temp coefficient Vishay axial resistors for under $1 with a temp coefficient down to a few ppm/C that you could base your secondary references on.
If you search the EEVblog for Hamon divider, you should be able to find links to Conrad's site where he explains it all.
Richard.
Edit: A quick Google search found Conrad's description of the Hamon Divider:
http://conradhoffman.com/HamonResistor.htmlConrad seems to be using a Hi-res meter, but all that is needed is a sensitive galvanometer to detect 0V +/- 1uV. Everything else can be done with resistors alone, so you do not need a 6 1/2 digit DMM to get the Hamon divider down in the 1ppm accuracy level.