marketmaker.ccPenyelidikan tentang sistem dagangan yang berfungsi dalam pasaran langsung.https://marketmaker.cc/msXGBoost for Return Direction: Class Imbalance and Decision Thresholdshttps://marketmaker.cc/ms/blog/xgboost-gradient-boosting-trading/https://marketmaker.cc/ms/blog/xgboost-gradient-boosting-trading/Return-direction classifiers are imbalanced problems, and 0.5 is the wrong decision threshold. Comparing scale_pos_weight, focal loss, and precision-constrained threshold optimization on crypto data - plus the engineering differences between XGBoost, LightGBM, and CatBoost.Tue, 25 Aug 2026 00:00:00 GMTmachine-learningxgboostgradient-boostingquantpredictionTransfer Entropy: Which Way Does Information Flow Between Crypto Assets?https://marketmaker.cc/ms/blog/transfer-entropy-crypto-information-flow/https://marketmaker.cc/ms/blog/transfer-entropy-crypto-information-flow/DCC-GARCH tells you when crypto dependence tightens. Transfer entropy tells you which way it points. A directed information-flow measure, its null calibration, and an honest account of what it does and does not add over average pairwise correlation.Mon, 24 Aug 2026 00:00:00 GMTcausal-inferencetransfer-entropyinformation-theorycryptonetworkTrade Classification When You Have No Side Flag: Tick, Quote, Lee-Ready, BVChttps://marketmaker.cc/ms/blog/trade-classification-algorithms/https://marketmaker.cc/ms/blog/trade-classification-algorithms/Reconstructing trade direction from OHLCV, bar history, or venues without an aggressor flag - the classic rules, the stale-quote problem, bulk volume classification, and how to measure them against crypto's free ground truth.Sun, 23 Aug 2026 00:00:00 GMTmicrostructuretrade-classificationLee-Readyorder-flowdeep-learningIrregular Time in Tick Models: Continuous-Time Encodings vs. Plain Positional Embeddingshttps://marketmaker.cc/ms/blog/tick-data-forecasting-deep-learning/https://marketmaker.cc/ms/blog/tick-data-forecasting-deep-learning/Sequence models fed tick data still assume regular spacing. Three ways to tell a Transformer when a tick actually happened - learnable-timescale continuous encoding, ODE-RNN latent state, and delta_t as a plain feature - and the ablation that decides between them.Sat, 22 Aug 2026 00:00:00 GMTHFTtick-datadeep-learninghigh-frequencypredictionSynthetic Control Methods for Evaluating Trading Strategieshttps://marketmaker.cc/ms/blog/synthetic-control-trading-evaluation/https://marketmaker.cc/ms/blog/synthetic-control-trading-evaluation/This series has priced the selection route to a false edge - DSR prices the winner, PBO prices the search. Neither touches confounding: the strategy that made money because volatility doubled the week you deployed it. The Synthetic Control Method builds a weighted counterfactual from a donor pool of untouched instruments and gives you a falsification criterion and a placebo p-value.Fri, 21 Aug 2026 00:00:00 GMTcausal-inferencesynthetic-controlcounterfactualevaluationeconometricsScoring Probabilistic Forecasts: CRPS, PIT Calibration, and DeepARhttps://marketmaker.cc/ms/blog/probabilistic-forecasting-trading/https://marketmaker.cc/ms/blog/probabilistic-forecasting-trading/How to evaluate a predictive distribution honestly - CRPS as a proper scoring rule, the PIT histogram as a calibration diagnostic, and DeepAR sampling in GluonTS.Thu, 20 Aug 2026 00:00:00 GMTforecastingprobabilisticquantile-regressionDeepARuncertaintyPhysics-Informed Neural Networks for Options Pricinghttps://marketmaker.cc/ms/blog/pinn-options-pricing/https://marketmaker.cc/ms/blog/pinn-options-pricing/Putting the Heston, free-boundary, and jump-diffusion pricing PDEs into a neural network loss - the log-price residual, the mixed-partial autograd trick, and what still needs measuring.Wed, 19 Aug 2026 00:00:00 GMTdeep-learningPINNoptionsBlack-ScholesPDEPCMCI: Causal Discovery in Multivariate Crypto Time Serieshttps://marketmaker.cc/ms/blog/pcmci-causal-discovery-crypto/https://marketmaker.cc/ms/blog/pcmci-causal-discovery-crypto/How PCMCI's two-stage MCI test recovers directed causal links between crypto assets where correlation and bivariate Granger cannot - the construction, the tigramite pipeline, and the real-data study it still needs.Tue, 18 Aug 2026 00:00:00 GMTcausal-inferencePCMCIcausal-discoverytime-seriescryptoOrder Flow Imbalance: The Cont-Kukanov-Stoikov Event Decompositionhttps://marketmaker.cc/ms/blog/order-flow-imbalance-prediction/https://marketmaker.cc/ms/blog/order-flow-imbalance-prediction/Turning raw book updates into a signed flow quantity: the CKS event decomposition, multi-level OFI with PCA reduction, and Lee-Ready trade classification - plus an honest accounting of what the headline R-squared actually measures.Mon, 17 Aug 2026 00:00:00 GMTmicrostructureorder-flowimbalancepredictionHFTNeural ODEs: Does Continuous Time Beat a Delta-Time Feature?https://marketmaker.cc/ms/blog/neural-ode-continuous-finance/https://marketmaker.cc/ms/blog/neural-ode-continuous-finance/Neural ODEs, Neural SDEs and continuous normalizing flows for irregularly-sampled market data - and the one ablation that decides whether continuous dynamics are worth their solver cost.Sun, 16 Aug 2026 00:00:00 GMTdeep-learningneural-ODEcontinuous-timeSDEdynamicsMulti-Task Learning for Simultaneous Price, Volume, and Volatility Predictionhttps://marketmaker.cc/ms/blog/multi-task-learning-trading/https://marketmaker.cc/ms/blog/multi-task-learning-trading/Does jointly predicting return, volume, and volatility actually help? Measuring loss-balancing schemes and diagnosing negative transfer through gradient cosine similarity - with a classical baseline and purged walk-forward folds.Sat, 15 Aug 2026 00:00:00 GMTdeep-learningmulti-taskshared-representationpredictionauxiliaryModel Pruning for Low-Latency Trading Inferencehttps://marketmaker.cc/ms/blog/model-pruning-low-latency-trading/https://marketmaker.cc/ms/blog/model-pruning-low-latency-trading/Magnitude and structured pruning, the Lottery Ticket Hypothesis, movement pruning, distillation and 2:4 sparsity - the methods behind shrinking a trading model, and what still has to be measured before any of it ships.Fri, 14 Aug 2026 00:00:00 GMTmodel-compressionpruninglottery-ticketlatencydeploymentUpdating the Volume Curve Intraday: Does Adaptive Forecasting Actually Help?https://marketmaker.cc/ms/blog/liquidity-prediction-execution/https://marketmaker.cc/ms/blog/liquidity-prediction-execution/Our VWAP article shipped a static, weekly-refit volume curve and called the forecaster the weakest link. This is the follow-up: a Bayesian intraday updater run against the same 500-parent BTCUSDT harness, with the IS delta conditioned on realized curve error.Thu, 13 Aug 2026 00:00:00 GMTmicrostructureliquidityexecutionpredictionorder-bookKoopman Operators and DMD: Do Market Modes Survive Out-of-Sample?https://marketmaker.cc/ms/blog/koopman-operator-market-dynamics/https://marketmaker.cc/ms/blog/koopman-operator-market-dynamics/Dynamic Mode Decomposition fits a linear operator to nonlinear market dynamics. The only question that matters: do the fitted modes persist from one window to the next, and does the rolling spectral radius lead realised volatility? Here is the measurement protocol and the code to run it.Wed, 12 Aug 2026 00:00:00 GMTmathematicsKoopmandynamical-systemsspectralpredictionKnowledge Distillation: Compressing Trading Models for Low-Latency Deploymenthttps://marketmaker.cc/ms/blog/knowledge-distillation-trading-models/https://marketmaker.cc/ms/blog/knowledge-distillation-trading-models/The blog's standing answer to the accuracy-vs-latency tension is a two-stage fast/slow split. Distillation is a different answer: train one small model to mimic the ensemble. The KD loss, temperature, born-again nets, early exits for a variable latency budget, and the distill-to-FPGA pipeline - plus the measurements that would decide whether it beats the two-stage split.Tue, 11 Aug 2026 00:00:00 GMTmodel-compressiondistillationlatencyHFTdeploymentThe Other Way a Regression Lies: Endogeneity, 2SLS, and the Gamma Calibration Problemhttps://marketmaker.cc/ms/blog/instrumental-variables-quantitative-finance/https://marketmaker.cc/ms/blog/instrumental-variables-quantitative-finance/Selection bias in the search is not the only way a regression fools you. When the regressor is correlated with the error, more data makes the estimate more confidently wrong. Instrumental variables applied to the one endogeneity problem this blog has already left open: permanent impact from net taker flow.Mon, 10 Aug 2026 00:00:00 GMTcausal-inferenceinstrumental-variables2SLSeconometricsendogeneityHawkes Processes for Order Arrival and Market Event Modelinghttps://marketmaker.cc/ms/blog/hawkes-process-order-arrival/https://marketmaker.cc/ms/blog/hawkes-process-order-arrival/Fitting a self-exciting point process to real crypto trade tape: where the three numbers (mu, alpha, beta) come from, how to estimate the branching ratio n, whether the exponential kernel survives a goodness-of-fit test, and how much n moves when you change the estimation window.Sun, 09 Aug 2026 00:00:00 GMTmicrostructureHawkes-processpoint-processorder-arrivalcryptoHFTHamiltonian Neural Networks: Does a Financial System Conserve Anything?https://marketmaker.cc/ms/blog/hamiltonian-neural-networks-finance/https://marketmaker.cc/ms/blog/hamiltonian-neural-networks-finance/Hamiltonian Neural Networks are provably stable - but stability is worthless if the conserved quantity does not exist. Learning a scalar H via autograd, symplectic integration, and the falsification test that decides whether a financial (q, p) pair is canonical at all.Sat, 08 Aug 2026 00:00:00 GMTdeep-learningHamiltonianphysics-informeddynamicsconservationToda-Yamamoto vs Differenced Granger: Does the BTC Lead-Lag Survive?https://marketmaker.cc/ms/blog/granger-causality-lead-lag-crypto/https://marketmaker.cc/ms/blog/granger-causality-lead-lag-crypto/Granger causality on crypto prices done two ways - differenced returns and Toda-Yamamoto on levels - with a correct Wald implementation, an effective-N corrected causality matrix, and a rolling-stability test of whether the lag is tradeable at all.Fri, 07 Aug 2026 00:00:00 GMTcausal-inferenceGranger-causalitylead-lagcryptoVARProses Gaussian untuk Pemodelan Harga Bukan Parametrikhttps://marketmaker.cc/ms/blog/gaussian-processes-trading/https://marketmaker.cc/ms/blog/gaussian-processes-trading/Reka bentuk kernel untuk siri masa kewangan - Kekasaran ibu, komposisi berkala setempat, campuran spektrum - dan kemungkinan kecil sebagai penyusun tetap yang tidak memerlukan set pengesahan. Selain itu senarai jujur ​​tentang perkara yang masih perlu diukur sebelum mana-mana daripadanya boleh didagangkan.Thu, 06 Aug 2026 00:00:00 GMTbayesiangaussian-processkerneluncertaintynon-parametricOperator Neural Fourier untuk Pemodelan Kewangan Berasaskan PDEhttps://marketmaker.cc/ms/blog/fourier-neural-operator-finance/https://marketmaker.cc/ms/blog/fourier-neural-operator-finance/Pembelajaran operator memetakan keseluruhan ruang fungsi, bukan titik - cara Operator Neural Fourier meparameterkan operator penyelesaian PDE dalam ruang frekuensi, perkara yang dibeli untuk penentuan harga opsyen, dan tuntutannya yang mana masih perlu diukur pada perkakasan sebenar.Wed, 05 Aug 2026 00:00:00 GMTdeep-learningFNOPDEoperator-learningoptionsAdakah Konteks Sepenuh Hari Mengatasi Sepuluh Minit? Perhatian Kilat dan Soalan Panjang Urutanhttps://marketmaker.cc/ms/blog/flash-attention-trading-models/https://marketmaker.cc/ms/blog/flash-attention-trading-models/Flash Attention menjadikan konteks dagangan 23,400 langkah bebas pengiraan - jubin, softmax dalam talian dan sempadan IO N^2 d^2 / M. Sama ada konteks yang lebih panjang itu menjadikan model itu lebih baik ialah soalan yang berasingan, dan ia perlu diukur, bukan ditegaskan.Tue, 04 Aug 2026 00:00:00 GMTdeep-learningattentionFlash-AttentionGPUoptimizationKaedah Ensembel: Menggabungkan Pelajar Lemah untuk Alfa Teguhhttps://marketmaker.cc/ms/blog/ensemble-methods-trading/https://marketmaker.cc/ms/blog/ensemble-methods-trading/Tiga perkara mengenai ensembel perdagangan yang blog ini tidak meliputi: mengapa istilah kovarians - bukan kiraan model - mendominasi ralat ensembel, cara tindanan berfungsi sebagai lapisan gabungan isyarat dengan ciri meta luar lipatan, dan sama ada jaringan pusing ganti merentas banyak isyarat sebenarnya mengurangkan kos pelaksanaan.Mon, 03 Aug 2026 00:00:00 GMTmachine-learningensemblebaggingstackingalphaDouble Machine Learning: Menganggar Parameter Sebab Bukkan Meramal Pulanganhttps://marketmaker.cc/ms/blog/double-ml-causal-trading/https://marketmaker.cc/ms/blog/double-ml-causal-trading/Semua model dalam blog ini setak ini menjawab 'apa meramalkan apa'. Double ML menjawab 'apa menyebabkan apa' - dengan ralat standard yang boleh anda pertahankan. Model separa linear, ortogonality Neyman, cross-fitting yang dibersihkan pada data buku pesanan, dan akaun jujur tentang mengapa selang keyakinan DML yang sah bertahan tepat untuk satu soalan pra-specified.Sun, 02 Aug 2026 00:00:00 GMTcausal-inferencedouble-MLtreatment-effecteconometricsquantKesan Rawak Heterogen untuk Dagangan menggunakan Causal Forestshttps://marketmaker.cc/ms/blog/causal-forests-treatment-effects-trading/https://marketmaker.cc/ms/blog/causal-forests-treatment-effects-trading/Setiap backtest dalam blog ini menganggar purata bersyarat. Causal forests menganggar kesan rawak bersyarat sebaliknya - tau(x) bukan mu(x) - dengan honest splitting, perwakilan berat kernel adaptif, dan ujian kalibrasi yang memberitahu anda sama ada heterogenitas yang anda jumpa adalah nyata.Sat, 01 Aug 2026 00:00:00 GMTcausal-inferencecausal-forestheterogeneous-effectseconometricstreatmentEpistemik vs Aleatorik: Mengukur Apa yang Model Pulangan Tidak Tahuhttps://marketmaker.cc/ms/blog/bayesian-neural-networks-trading/https://marketmaker.cc/ms/blog/bayesian-neural-networks-trading/Setiap peraturan saiz kedudukan dalam blog ini meruntuhkan ketidakpastian kepada satu skalar. MC Dropout dan ansambl dalam membahagikannya kepada kejahilan model dan hingar pasaran - dan kedua-duanya memerlukan saiz kedudukan yang berbeza.Fri, 31 Jul 2026 00:00:00 GMTbayesianuncertaintyneural-networkriskposition-sizingAutoML untuk Saluran Paip Dagangan Sistematikhttps://marketmaker.cc/ms/blog/automl-systematic-trading/https://marketmaker.cc/ms/blog/automl-systematic-trading/Automated feature generation (tsfresh, Featuretools), budget-aware model search (FLAML's cost-frugal optimizer), and the WorldQuant formulaic alpha factory - the parts of the research pipeline this blog's search-and-overfit arc never covered, and what the arc's own results say about them.Thu, 30 Jul 2026 00:00:00 GMTAutoMLNASfeature-engineeringautomationquantLiquidation cascades as a trading signal: reading forced, pre-announced flowhttps://marketmaker.cc/ms/blog/liquidation-cascades-trading-signal/https://marketmaker.cc/ms/blog/liquidation-cascades-trading-signal/Every leveraged position advertises the price at which it must be sold. How to build the on-chain liquidation depth chart and the CEX liquidation heatmap, model cascade dynamics as a reproduction number, and trade forced flow as a signal instead of only fearing it as a risk.Wed, 29 Jul 2026 00:00:00 GMTliquidationscascadetrading signaldefiperpetualsorder bookmarket microstructurecryptoActive Uniswap v3 LPing as Market Making: Range Selection, Rebalancing, and Delta Hedginghttps://marketmaker.cc/ms/blog/uniswap-v3-lp-strategies-hedging/https://marketmaker.cc/ms/blog/uniswap-v3-lp-strategies-hedging/An active v3 LP is running a market-making book with gas costs and no cancel button. Size ranges from a GARCH vol forecast, frame rebalancing as fee-gain vs realized-cost, derive the position delta from tick math and hedge it with perps, and understand JIT liquidity and the honest pitfalls of backtesting LP from on-chain data.Tue, 28 Jul 2026 00:00:00 GMTuniswapdefimarket makingdelta hedginglvrjit liquidityquantethereumThe MEV supply chain: PBS, MEV-Boost, and who actually captures the valuehttps://marketmaker.cc/ms/blog/mev-supply-chain-pbs-mevboost/https://marketmaker.cc/ms/blog/mev-supply-chain-pbs-mevboost/MEV stopped being a lone bot and became an assembly line: searcher, builder, relay, proposer. A technical walk through proposer-builder separation, MEV-Boost's sealed-bid block auction, why the searcher's margin gets bid away, order-flow auctions as the new moat, and how Solana's Jito model differs.Mon, 27 Jul 2026 00:00:00 GMTmevpbsmev-boostflashbotsjitoorder flow auctiondefiethereumArbitraj dalam rantaian: kitaran atom, pinjaman kilat, dan lelongan yang perlu anda menangihttps://marketmaker.cc/ms/blog/onchain-arbitrage-atomic-flash-loans/https://marketmaker.cc/ms/blog/onchain-arbitrage-atomic-flash-loans/Mencari kitaran arbitraj dalam rantaian adalah separuh yang mudah. Separuh yang sukar adalah memenangi lelongan keutamaan bida tertutup menentang setiap pencari lain yang menemui kitaran yang sama. Pecahan teknikal atomicity, pinjaman kilat, backrunning, dan di mana keuntungan arb sebenarnya bertahan - termasuk CEX-DEX bukan atom.Sun, 26 Jul 2026 00:00:00 GMTarbitrageflash loansmevdefibackrunningcex-dexethereumatomicityImpermanent Loss dan LVR: matematik sebenar keuntungan LPhttps://marketmaker.cc/ms/blog/impermanent-loss-lvr-lp-profitability/https://marketmaker.cc/ms/blog/impermanent-loss-lvr-lp-profitability/Impermanent loss, leverage kecairan tertumpu, dan loss-versus-rebalancing yang diperoleh daripada prinsip pertama. Bentuk tertutup, hasil σ²/8, markout per-blok, dan syarat yuran-berbanding-volatiliti yang tepat di mana LP sebenarnya mengalahkan pegangan.Sat, 25 Jul 2026 00:00:00 GMTdefiammimpermanent losslvruniswapmarket makingvolatilityquantDi dalam slice: taktik child-order antara scheduler anda dan exchangehttps://marketmaker.cc/ms/blog/child-order-execution-tactics/https://marketmaker.cc/ms/blog/child-order-execution-tactics/Almgren-Chriss dan VWAP hanya menetapkan bajet slice - eksekusi dimenangi pada lapisan taktik. Timer eskalasi, matematik break-even maker-taker, semantik queue amend vs cancel-replace di Binance/OKX/CME, anti-signaling iceberg, dan state machine Python per-slice.Fri, 24 Jul 2026 00:00:00 GMTexecutionchild ordersorder tacticsqueue positionmaker-takericeberg ordersmicrostructurepythonLengkung slippage, bukan pemalar slippage: model kos yang bertahan apabila bersentuhan dengan dagangan langsunghttps://marketmaker.cc/ms/blog/slippage-cost-models-backtest/https://marketmaker.cc/ms/blog/slippage-cost-models-backtest/Ganti slippage bps tetap dengan lengkung kos yang bergantung pada saiz, volatiliti dan kecairan: hukum punca kuasa dua, pemasangan lengkung daripada fill TCA atau data awam, pengganda tekanan rejim, dan sebab papan pendahulu strategi anda tersusun semula.Thu, 23 Jul 2026 00:00:00 GMTslippagekos transaksiimpak pasaranbacktesteksekusitcahukum punca kuasa duamodel kosPenghalaan Pesanan Pintar dalam Kripto: Satu Pesanan, Dua Belas Venue, Tiada NBBOhttps://marketmaker.cc/ms/blog/smart-order-routing-crypto/https://marketmaker.cc/ms/blog/smart-order-routing-crypto/Kenapa SOR kripto lebih sukar berbanding penghalaan ekuiti: tiada tape konsolidasi, kecairan hantu, modal pra-dana. Satu pengoptimuman penghalaan cembung lengkap dengan matematik dan Python, taktik sedar-maker, dan jadual liga markout mengikut venue daripada TCA anda sendiri.Wed, 22 Jul 2026 00:00:00 GMTeksekusipenghalaan pesanan pintarmikrostruktur pasaranpemecahan kecairankriptokuantitatifpythontcaKeputusan maker-taker: tingkat fi, rebat, dan kos sebenar melintasi spreadhttps://marketmaker.cc/ms/blog/maker-taker-fees-rebates-execution/https://marketmaker.cc/ms/blog/maker-taker-fees-rebates-execution/Maker vs taker ialah dagangan adverse selection yang berselindung di sebalik jadual fi. Matematik titik pulang modal melalui Glosten-Milgrom, tingkat fi kripto sebenar dan diskaun token, sejarah rebate mining, nilai rebat yang bergantung pada queue, dan pemodelan tingkat fi dalam backtest.Tue, 21 Jul 2026 00:00:00 GMTmaker takerfi daganganrebatadverse selectionmarket makingeksekusimarket microstructurebacktestImplementation shortfall dan DIY TCA: mengukur kos sebenar eksekusi andahttps://marketmaker.cc/ms/blog/implementation-shortfall-tca-execution/https://marketmaker.cc/ms/blog/implementation-shortfall-tca-execution/Implementation shortfall Perold diubah menjadi saluran paip TCA yang berfungsi untuk bot kripto: penguraian arrival-price, lengkung markout pada t+1s/10s/60s, ~200 baris Python ke atas fill anda sendiri, dan memasukkan semula hasilnya ke dalam model kos backtest anda.Mon, 20 Jul 2026 00:00:00 GMTexecutiontcaimplementation-shortfallmarkoutskos-transaksialgotradingbacktestmarket-microstructureLiquidasi on-chain: mekanik Aave dan Compound, dan perniagaan bot di sekitarnyahttps://marketmaker.cc/ms/blog/onchain-liquidations-aave-compound/https://marketmaker.cc/ms/blog/onchain-liquidations-aave-compound/Bagaimana liquidasi Aave v3 dan Compound sebenarnya berfungsi: matematik health factor, close factor, liquidation bonus, pencetus oracle Chainlink, dan apa yang diperlukan untuk menjalankan bot liquidasi flash-loan pada 2026.Sun, 19 Jul 2026 00:00:00 GMTdefiliquidasiaavecompoundmevchainlinkflash loanscryptoUniswap v3 untuk Quant: Kecairan Tertumpu dan Matematik Tick dari Prinsip Asashttps://marketmaker.cc/ms/blog/uniswap-v3-concentrated-liquidity-quants/https://marketmaker.cc/ms/blog/uniswap-v3-concentrated-liquidity-quants/Bagaimana Uniswap v3 sebenarnya berfungsi di sebalik tabir: rizab maya, kecairan L, sqrtPriceX96, matematik tick, feeGrowthInside, dan mengapa kedudukan LP adalah bayaran pendek-volatiliti. Lengkap dengan formula tepat dan contoh berangka yang diselesaikan.Sat, 18 Jul 2026 00:00:00 GMTuniswapdefiammkecairan tertumpupembuatan pasaranlvrkuantethereumAnatomi MEV: sandwich, frontrunning, dan hutan gelap mempoolhttps://marketmaker.cc/ms/blog/mev-sandwich-frontrunning-mempool/https://marketmaker.cc/ms/blog/mev-sandwich-frontrunning-mempool/Setiap swap yang anda hantar ke RPC awam ialah order had yang kelihatan oleh semua tetapi tiada siapa wajib memenuhinya. Pecahan teknikal tentang cara pengekstrakan MEV berfungsi: penyusunan mempool, matematik sandwich pada Uniswap v2, PGA, Flashbots, dan pertahanan pedagang.Fri, 17 Jul 2026 00:00:00 GMTmevsandwich attackfrontrunningmempoolflashbotsdefiethereumSimulasi fill: tangga dari fantasi harga penutup kepada realiti sedar-queuehttps://marketmaker.cc/ms/blog/fill-simulation-partial-fills-backtest/https://marketmaker.cc/ms/blog/fill-simulation-partial-fills-backtest/Lima anak tangga ketepatan simulasi fill - dari fill harga penutup kepada model kebarangkalian kedudukan queue. Partial fill sebagai state machine, sempadan kebarangkalian fill limit sebagai bracket PnL, dan gelung kalibrasi terhadap fill live.Thu, 16 Jul 2026 00:00:00 GMTsimulasi fillbacktestlimit orderkedudukan queuepartial fillmarket makingeksekusimarket microstructureTWAP lawan VWAP lawan POV: memilih penanda aras eksekusi (dan mengetahui bila setiap satu menipu anda)https://marketmaker.cc/ms/blog/twap-vwap-pov-execution-algorithms/https://marketmaker.cc/ms/blog/twap-vwap-pov-execution-algorithms/TWAP, VWAP dan POV adalah pertaruhan ke atas ramalan volum. Kami membedah andaian tersembunyi setiap penjadual, membina lengkung volum intraharian kripto, dan menjalankan ketiga-tiganya secara head-to-head ke atas data L2 yang dimainkan semula.Wed, 15 Jul 2026 00:00:00 GMTexecutiontwapvwappovimplementation-shortfallvolume-curvealgotradingbacktestAlmgren-Chriss Tanpa Berselindung: Pelaksanaan Optimum yang Boleh Anda Laksanakan dalam Satu Petanghttps://marketmaker.cc/ms/blog/almgren-chriss-optimal-execution/https://marketmaker.cc/ms/blog/almgren-chriss-optimal-execution/Terbitan penuh model pelaksanaan optimum Almgren-Chriss: kesan linear, trajektori sinh/cosh, sempadan cekap, dan kod Python yang berfungsi untuk menentukur eta, gamma dan sigma daripada data L2 dan dagangan Binance.Tue, 14 Jul 2026 00:00:00 GMTpelaksanaanalmgren-chrisskesan pasaranpelaksanaan optimumtwapquantpythonkriptoPenyasaran Kemeruapan dan Berdagang dengan Ramalan GARCHhttps://marketmaker.cc/ms/blog/volatility-targeting-garch-strategy/https://marketmaker.cc/ms/blog/volatility-targeting-garch-strategy/Ramalan kemeruapan GARCH hanya bernilai jika ia menambah baik keputusan dagangan. Kami membina strategi kripto bersasarkan kemeruapan, menilai kualiti ramalan secara jujur, dan membandingkan GARCH dengan garis dasar realized-vol dan EWMA dalam backtest walk-forward.Mon, 13 Jul 2026 00:00:00 GMTvolatilityGARCHvolatility-targetingbacktestingriskcryptoalgorithmic-tradingDCC-GARCH: Korelasi Dinamik untuk Pairs dan Risiko Portfoliohttps://marketmaker.cc/ms/blog/dcc-garch-dynamic-correlation-crypto/https://marketmaker.cc/ms/blog/dcc-garch-dynamic-correlation-crypto/Korelasi kripto tidak konstan - ia melonjak ke arah 1 dalam setiap drawdown. DCC-GARCH memodelkan matriks korelasi yang berubah mengikut masa, memberikan anda nisbah lindung nilai dinamik untuk pairs trading dan risiko portfolio yang jujur dan berubah mengikut masa.Sun, 12 Jul 2026 00:00:00 GMTvolatilityGARCHDCCcorrelationportfoliopairs-tradingcryptoGARCH Tak Simetri dan Ekor Tebal: EGARCH, GJR, dan Student-thttps://marketmaker.cc/ms/blog/asymmetric-garch-crypto-leverage/https://marketmaker.cc/ms/blog/asymmetric-garch-crypto-leverage/GARCH(1,1) biasa melayan berita baik dan buruk secara sama serta menganggap kejutan Gaussian. EGARCH, GJR-GARCH, dan inovasi Student-t/skew-t membetulkan kedua-duanya - dan memberikan anda VaR serta Expected Shortfall yang jujur untuk kripto.Sat, 11 Jul 2026 00:00:00 GMTvolatilityGARCHEGARCHriskVaRcryptoalgorithmic-tradingGARCH(1,1): Meramal Volatiliti Kriptohttps://marketmaker.cc/ms/blog/garch-volatility-forecasting-crypto/https://marketmaker.cc/ms/blog/garch-volatility-forecasting-crypto/Bagaimana model GARCH(1,1) menangkap pengelompokan volatiliti dalam kripto, cara memasangnya melalui kemungkinan maksimum dengan pustaka arch, dan cara menukar ramalan varians bersyarat menjadi saiz kedudukan serta stop dinamik.Fri, 10 Jul 2026 00:00:00 GMTvolatilityGARCHriskforecastingcryptoalgorithmic-tradingYang Negatif dengan Jujur: Puluhan Ribu Backtest, Lima Mata Wang Utama, Tiada Kelebihan yang Mantaphttps://marketmaker.cc/ms/blog/honest-negative-no-robust-edge/https://marketmaker.cc/ms/blog/honest-negative-no-robust-edge/Kemuncak babak pencarian-dan-overfit, dan ia berakhir dengan satu keputusan negatif - yang betul. Satu carian dual-timeframe simbol tunggal pada ETHUSDT menemui satu konfigurasi bernilai +16.35% out-of-sample dan +2.62% pada satu holdout yang tidak disentuh; Deflated Sharpe Ratio, dengan mengambil kira ~37,000 percubaan, mendeflasikannya kepada 0.00. Satu larian merentas-instrumen ke atas lima mata wang utama (ETH/BTC/SOL/BNB/XRP, ~1.18M bar 1m setiap satu), memilih mengikut median out-of-sample, membunuhnya buat selama-lamanya: dual DSR 0.24 / PBO 0.264, triple DSR 0.14 / PBO 0.327 - kedua-duanya gagal gate. Juara ini menguntungkan pada 1 daripada 5 simbol dan negatif pada selebihnya. Inilah tujuan radas anti-overfit: untuk menghalang anda daripada melancarkan yang terbaik daripada hingar sebagai alpha.Thu, 09 Jul 2026 00:00:00 GMTalgotradingbacktestoverfittingdeflated-sharpepbokeputusan-negatifpengesahanMembuktikan Tiada Look-Ahead dalam Ujian Balik Multi-Timeframe: Ganggu Masa Depan, Buktikan Masa Lalu Tidak Dapat Melihatnyahttps://marketmaker.cc/ms/blog/proving-no-lookahead-multi-timeframe/https://marketmaker.cc/ms/blog/proving-no-lookahead-multi-timeframe/Ujian balik (backtest) multi-timeframe membocorkan masa depan melalui bar timeframe lebih tinggi yang sedang terbentuk, yang harga penutupan akhirnya belum wujud lagi. Anda tidak boleh mencapai keyakinan hanya dengan menyemak kod - ia perlu diuji. Kami mengulangi semula peraturan bar-tertutup bot langsung dengan tepat, kemudian membuktikan tiada kebocoran dengan ujian anjakan-masa-depan: ganggu setiap bar masa depan dan sahkan setiap isyarat serta dagangan masa lalu kekal sama tepat sehingga bit terakhir. 25/25 semakan pariti, dan ujian ini benar-benar berkesan.Wed, 08 Jul 2026 00:00:00 GMTalgotradingbacktestbias look-aheadmulti-timeframekebocoran datapengesahanApabila GPU Berbaloi: Roofline Sapuan Parameter, di Mana Tajuk Utama 167x Sebenarnya 27x Algoritma Darab 6.2x Perkakasanhttps://marketmaker.cc/ms/blog/when-gpu-pays-off-sweep-roofline/https://marketmaker.cc/ms/blog/when-gpu-pays-off-sweep-roofline/Kelebihan GPU berbanding CPU tumbuh dengan saiz kelompok - 54.5x pada satu kombo setiap panggilan hingga 359.6x pada 61 dalam prapengiraan penunjuk pelbagai jangka masa kami - kerana sapuan kecil tidak dapat mengamortisasi overhead pelancaran kernel dan pemindahan. Kami menguraikan tajuk utama 167x kepada kemenangan algoritma 27x yang turut membantu CPU dan kemenangan perkakasan 6.2x, menunjukkan kelebihan sebenar GPU-berbanding-CPU-terbaik hanyalah 3.2x untuk jangka masa tunggal dan 6.2x untuk pelbagai jangka masa, dan memberikan panduan keputusan tentang berapa lebar sapuan perlu sebelum GPU berbaloi dibeli.Tue, 07 Jul 2026 00:00:00 GMTalgotradingbacktestprestasigpurooflinepengoptimuman