Why is this line of the Caro-Kann playable for white?

Question

Today in an online game my opponent opened with e4 and I played the Caro-Kann, as I usually do. We went into a line that was new to me, shown below:

rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1 
1. e4 c6 2. Nf3 d5 3. d3 dxe4 4. dxe4 Qxd1+ 5. Kxd1

I felt pretty good about this position, since white gave up their castling rights for a small advantage in development, which doesn't look worrying at all with the queens already traded off and I went on to win the game. Much to my surprise however, I later discovered that after 5. Kxd1 Stockfish slightly prefers white's position, and that this line has been played by various superGMs with good results.

Why is the position after 5. Kxd1 good for white? What is white's compensation for being unable to castle? What are the main ideas for both sides in this position?

Answer 1

What is white's compensation for being unable to castle?

With queens off the board castling rights are not that important.

Why is the position after 5. Kxd1 good for white?

White has a center pawn and a nicely developed knight while black has a pawn on c6 taking away the b knight's best square, no center pawn and no pieces developed. In these open positions fast development is very important.

Answer 2

I have been playing this variation for many years, I think it is excellent for White. It gives a stable advantage and White does not risk anything.

Why is the endgame so good? The short answer is, White will play a4 and if possible a5 at some point and point out the compromised Black structure on the queen side. This might sound surprising but if you play a couple of games you will realize that this is difficult to combat. A typical line might go something like this

rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1 
1. e4 c6 2. Nf3 d5 3. d3 dxe4 4. dxe4 Qxd1+ 5. Kxd1 Nf6 6. Nbd2 Bg4 7. Be2 Nbd7 8. Ne1 Bxe2+ 9. Kxe2 e6 10. a4

and White is already much better. You continue with a5 if you can, the White knight goes to d3 and the other to c4 if allowed. White will play f3 to completely solidify the center and stay with a good Bishop. Usually that Bishop goes to e3 putting pressure on a7 which holds Black's rook back. If Black goes a6 (or a5 immediately here) there's a big weakness on b6. Then White simply puts the rooks on the d file and slowly advances.

The computer even gives the option to play h4 at some point and push on the king side too though I haven't been able to make this work in a human game consistently so I would probably not recommend it.

What can Black do to combat this? To be honest, I don't know. If I were a Caro Kann player I would simply not go into this endgame. With precise play you will hold but it's no fun and the winning chances are near nil. 3...Bg4 can lead to some sharp variations where White plays h3 g4 and Ne5 at some point, following up with h4 and it can become very tactical. If you like a tactical mess you can play that with Black. 3...g6 is more solid and will likely transpose into some hyper accelerated Dragon Sicilian after 4.e5 c5 5.c3 followed by d4 or similar variations. The alternative is 3...Qc7 which fights for the e5 square more directly and "threatens" that e5 move.