First, I'd not rely too completely on the whole Stevens' schema. It is meant to be a guide, not a prison. I wrote an article about this on Medium.
Second, why do you think ratio scales can't be negative? The key thing about a ratio scale is that ratios make sense. But I see no reason you cannot have a ratio with a negative number. For interval level data, ratios make no sense. E.g. temperature in Celsius or Fahrenheit is interval because e.g. 100 degrees (on either scale) is not twice as hot as 50 degrees (on the same scale). This can be seen by comparing C and F. 100 C = 212 F, but 50 C = 122 F (not 106). But a displacement of -1 meter is -1 the displacement of +1 meter.
Third, Stevens originally talked about permissible transformations. For interval data, we can transform in any way that preserves the relative size of the intervals. For ratio data, we may only multiply. See my article for more explanation and examples.