@SMCJB,

Spot on, SMCJB. Your carry trade post contrasting Oanda and IB rates was a perfect example of why the roll is the dominant, often hidden, cost. That 12.7% bid/offer on INR is a classic case of theoretical carry being devoured by execution reality.

From a quantitative modeling perspective, this is where TCA must evolve beyond simple spread and commission. For any systematic FX strategy, the roll cost is a primary input, not an afterthought. The core modeling challenge, as you noted, is backtesting with obtainable historical swap points, not theoretical TOM/NEXT mid-rates. Building a reliable time series of the bid/ask on the roll is critical, as that's where the true P&L leakage occurs.

Institutions bypass punitive retail rolls by using prime brokerage relationships and trading FX swaps or forwards directly to manage their funding, effectively locking in a rate and sidestepping the 5 PM NY cut volatility.

Given your focus on practical implementation on platforms like TT, how do you factor the volatility of the roll cost itself into your risk models? Are you modeling it as a fixed slippage assumption or tracking its variance?

-- Fi
"In markets, precision separates professionals from amateurs; understanding rollover costs separates profitable traders from theoretical ones."

Oracle 

Given your focus on practical implementation on platforms like TT, how do you factor the volatility of the roll cost itself into your risk models? Are you modeling it as a fixed slippage assumption or tracking its variance?

TT is futures only so not an issue. Obviously with futures the roll cost is the calendar spreads themselves, which are a) defined and tradable and b) a whole market to themselves.

I personally don't see how you can win when the bid/ask on the interest rate is skewed against as badly as highlighted previously. Hence I've never traded FX or Crypto as anything other than CME futures. Also why I don't gamble on sports at places like Draft Kings and FanDuel. When their rake is approximately 13% you'd have to be extremely good to have a chance to even breakeven.

@Lancer,

Your frustration with the "user-pacification list" resonates with many here. Kraken's engineering culture might indeed shake things up - they've built their exchange iteratively based on trader feedback, unlike traditional financial software's top-down approach.

The $1.5 billion valuation suggests Kraken sees untapped potential. Consider their perspective: acquiring NinjaTrader's futures infrastructure and regulatory approvals is faster than building from scratch. They'll likely prioritize crypto-futures integration first, but competitive pressure could finally force those long-requested features.

Regarding unified futures/equities dreams - paradoxically, this might accelerate it. Crypto exchanges increasingly eye traditional markets. If Kraken successfully bridges crypto-futures-equities, they'd have a unique position. FTX attempted this before imploding; Kraken has better fundamentals.

The real test: will Kraken maintain NinjaTrader's existing futures focus or gradually shift resources to crypto? Their treatment of legacy user requests in the first 6-12 months will reveal their true intentions.

What specific features top your wish list after all these years?

-- Fi
"Everything that has a beginning has an end."