Quote from darkhorse:
"With many market strategies, risk goes down as leverage goes up"
I would argue that statement is inaccurate in a way that is quite important.
All things being equal, leverage should not have impact on the safeness or riskiness of a strategy. Unless you use too much of it, in which case risk of ruin becomes unacceptable.
The purpose of leverage is to amplify a potential return, at the cost of also amplifying potential drawdowns. If an unlevered strategy provides a 10% return and 3% historical drawdowns, then trading it at 10X would create a profile of 100% return and 30% drawdowns, or what have you.
Point being here, leverage is an amplifier in both directions. Availability of leverage should not change the risk / return profile of the strategy itself, independent of "tap out" considerations (how much money you can lose on the downside before being wiped out).
Of course, real world factors make the leverage question interesting. What Long Term Capital Management did was essentially argue their "hoovering up nickels" strategy had a drawdown risk of zero, allowing them to take a 50 basis point return (or whatever it was) and lever it to ridiculous orders of magnitude. We know how that turned out.
Anyway, I think what you are trying to say is that a certain amount of leverage is needed in order for XYZ market neutral strategy to be implemented properly.
But what you are really saying is that, a certain amount of AMPLIFICATION is needed to make the market neutral RETURNS worth gunning for in the first place, because unlevered market neutral is just not worth doing.
Market neutral is about units of risk in a certain ratio to each other. So if you have X units of risk bullish, you want to be sure you can also have X units of risk bearish, and so on. (A gross oversimplification, but useful for the purpose of this explanation.)
The thing is, it is possible to run a market neutral strategy in a completely unlevered account. You would just have to trade very, very small. This would not impact the ability to EXECUTE the strategy properly... but it WOULD impact the return.
So I would dispute your argument that "higher leverage results in a higher sharpe ratio and smaller drawdowns."
What you are really saying (I think) is that to run XYZ strategy at X size, to produce X return profile, a certain amount of leverage elbow room is required, and without that elbow room you are constrained. But the logical conclusion from this is not that more leverage = safer execution. It is to ask the following questions:
* What are my realistic returns for running this strategy unlevered
* How much do I want to amplify those returns with leverage, given the trade-off of amplified drawdowns as well
Keeping in mind, too, Michael Milken's post-crisis observation that "leverage is not a business model." Meaning, if a strategy produces tiny returns on an unlevered basis, the light may not be worth the candle in terms of levering it up significantly.
So all in all, it is not correct to say that increased leverage lowers risk. If one has a strategy in which more leverage is needed just to implement the strat properly, then the initial position sizes were too big in the first place. And if reducing position size to properly fit the available capital also reduces the return below the cost of funding, well, that is an issue with the strategy itself.