With all respect to Roach: "Even a broken clock is right twice a day"
Personally I have yet to encounter an economist that has been able to preddict market turns with a more than 50% probability. Most I know and follow (although honestly I don't spend much time listening to economists anymore) enjoy even less than a 50% success rate.
Personally, I agree with Lynch's comments: "I only spend 15 minutes a year trying to predict the economy".